View
1
Download
0
Category
Preview:
Citation preview
archaeological 3Jn0t(tute of
AMERICAN
JOURNAL OF ARCHAEOLOGY
THE JOURNAL OF THE ARCHAEOLOGICAL
INSTITUTE OF AMERICA
VOLUME XYI
1912
NORWOOD, MASS.
PUBLISHED FOR THE INSTITUTE BY
Wqt Nortoootr ^regs
NEW YORK: THE MACMILLAN COMPANY
LONDON: MACMILLAN & CO., LTD.
/3
Archaeological
Institute of
America
WHO BUILT THE ARCH OF CONSTANTINE ? ITS
HISTORY FROM DOMITIAN TO CONSTANTINE
HARDLY anything might seem more audacious than to
deny that the arch of Constantine was built in honor of that
emperor; yet the really amazing thing is our failure to attend
to the numerous hints that this arch had existed long before
Constantine. Artists and archaeologists have always been un-
able to explain how an architect of the decadent age of Con-
stantine could have given to this arch its marvellous proportionsand silhouette, which set it above all other arches, even those
of the golden age (Fig. 1). Historians have been puzzled by the
silence of that early catalogue of the buildings at Rome, the Noti-
tia, issued before Constantine's death (334 A. D.), which assigns to
Constantine, apparently, only the Janus in the Forum Boanum.The same Notitia increases the mystery by speaking of an Arcus
Novus on the Via Lata, which can only be the arch of Diocle-
tian, dedicated in 303. If in 334 the arch of 303 was still the
latest of triumphal arches, how could an arch have been built
to Constantine in 315 ? Besides, a student of Roman law
would argue that it was against the unbroken tenets of tradi-
tion and law to erect such an arch to an emperor who had not
actually been decreed a triumph and whose victories had been
not over a foreign but over a domestic foe. According to
ancient literature and law, therefore, there was not and could
not have been a triumphal arch of Constantine, in the sense
that it was built expressly for Constantine. It is quite differ-
ent if the arch could be recognized as an already existing arch
rededicated in his honor.
It is my expectation to proye in this paper that the arch was
built long before Constantine ; also to show that its construc-
American Journal of Archaeology, Second Series. Journal of the oaoArchaeological Institute of America, Vol. XVI (1912), No. 3.
THE ARCH OF CONSTANTINE 369
tion should probably be ascribed to the Emperor Domitian,shortly before or after 90 A.D., some 225 years before the dedi-cation to Constantine. After the assassination of Domitian,his memoriae damnatio by the senate condemned to mutilation,'and sometimes to destruction, all his public monuments, andespecially his memorial and triumphal arches, which wereclosest to him, personally. The dedicatory inscriptions, thestatues and reliefs in his honor, were destroyed. His works
FIGURE 1. THE ARCH OF CONSTANTINE, SEEN FROM THE NORTH.
where spared became ownerless and could be rededicated by or
to any emperor, as was the case, for instance, with the Forumof Nerva. But, throughout the second century, this arch, so
strongly associated with the odious memory of a tyrant', re-
mained unchanged arid unclaimed, for during this prosperous
age of the Antonines the senate continued to build special
arches for each triumphing emperor. It was only during the
third century, when Rome, impoverished and suffering from
the frequent absence of the emperors, with an art in constant
370 A. L. FROTHINGHAM
decay, and with building operations almost suspended for a
half century, resorted to makeshifts in the way of triumphalmonuments. Between 203 when the senate built the arch to
Septimius Severus and 303 when one was consecrated to Dio-
cletian, we know of the erection of but a single triumphal arch,
that of Gordian III, ca. 240. What was done by the senate
during these hundred years to commemorate imperial victories ?
I expect to show that the senate utilized for this purpose the
ex-Domitianic arch, turning this wound-scarred war-horse into
a marvellous historic bulletin board, a triumphal mosaic and
palimpsest, which became the quintessence of Roman history
during the third century. Then, between 312 and 315, after
it had thus long been purged of its original evil association
and, as its inscription boasts, become " famous for its manytriumphs," its evolution closed, and it was once more dedicated
to a single emperor, to Constantine, after a unique and varied
career, to be honored throughout the ages as a monument to
the first Christian emperor.It has been universally believed,
1 on the apparently unim-
peachable authority of the dedicatory inscription on the arch,
and on that authority alone, that when the Romans, grateful to
Constantine for reestablishing peace after his victory over
Maxentius, just outside Rome, in 312, decided to commemoratethe event by a triumphal arch, the architect gathered from
several earlier monuments a number of bas-reliefs, statues, and
architectural members, especially the main cornice, columns,
and pilasters, and built all this material into the fabric of the
arch as he erected it. To these spoils he is supposed to have
added whatever was needed to complete the design, by the
handiwork of contemporary artists, in the decadent style of the
Constantinian age. Until quite recently it was supposed that
the earlier sculptures that were so used were all of the time of
Trajan and taken from one of his arches either that on the
Via 'Appia or that in the Forum of Trajan or from someother part of his forum. But this theory, due to the current
ignorance of the historic phases of Roman sculpture, was
1 The Bibliography of the arch is too voluminous to be given here, and it
would be superfluous. Good lists are given by Mile. Bieber, by Sieveking, byArndt, and by other authors of the studies quoted in the following notes.
THE ARCH OF CONSTANTINE 371
shattered in 1889 and 1890 by Petersen,1 who showed that the
eight large reliefs of the attic belonged originally to a
triumphal arch of Marcus Aurelius, and who also proposed a
new interpretation of the eight medallions. It was suggestedthat the main cornice with its pilasters and columns, whichwere too beautiful to be Constantinian, were taken from the
same arch of Marcus Aurelius, together with the statues of
barbarians on the attic. Some years later,2 a new impetus came
from a detailed study of the eight medallions in the central
zone, which led Arndt to attribute these exquisite works not
to the Trajanic age, but to the neo-Hellenic art of Hadrian.
An English critic 3 then put forward the suggestion that theywere of the earlier Flavian age, were in fact Domitianic, taken
from the Domus gentis Flaviae. Almost at once, a German
archaeologist, Sieveking,4 while accepting the Flavian date for
four of the medallions, saw in the other four the art of Hadrian.
Then, quite recently, the publication on a large scale, from
casts,5 of the heads in the medallions has led to an interesting
discussion in which a number of critics have taken part, and
in the course of which Sieveking6 withdrew his dual suggestion
and joined those who believe in the Hadrianic theory. It has
been supposed that in these medallions, as elsewhere, the
original head of the emperor was changed into a portrait of
1 ' I rilievi tondi dell'Arco di Costantino,' in Rom. Mitt. 1889, p. 314;and
'Die Attikareliefs am Constantinsbogen,' ibid. 1890, p. 73. Cf. article byMonaci in B. Com. Rom. 1900, p. 25 ff. Later study by Petersen in Neue
Jahrb. f. Rlass. Alt. 1906, p. 522 ff.
2 Arndt, in Denk. griech. u. rdm. Skulptur, text to pis. 555, 559, 560, 565.
3 Stuart Jones,' Notes on Roman Historical Sculptures
' in B.8.R. III, p. 213,
published in 1905.
< 'Die Medaillons am Konstantinsbogen,1 Rom. Mitt. XXII, 1906, p. 345 ff.
5 By Salomon Reinach in Revue Archeologique, XVII, 1911, pis. I-XVII,
with interesting symposium of opinions by S. de Ricci, Studniczka, and others.
Cf. Revue Arch. XVII, 1911, p. 465.
6 Berl. Phil W. 1911, No. 39. The article which caused his reversal of
opinion was one on the medallions by Mile. Bieber (Rom. Mitt. 19J1, p. 274),
which illustrates the danger of basing a study as delicate, as aesthetic, and as
detailed as that of Sieveking on an examination of mere photographs instead of
the monument itself. It was a result which I predicted to Dr. Htilsen when he
received Dr. Sieveking's article for publication. Such facile criticism without
investigation of the originals ought to be discouraged, as it tends to confuse and
lower archaeological standards.
372 A. L. FROTHINGIIAM
Constantine when they were used on the arch ; but as certain
imperial heads were worked over to represent not Constantine,
but some emperor or emperors of about the middle of the third
century or later, critics suggested the names of Claudius
Gothicus, Philip, Carus, Carinus, and even of Constantino's
father, Constantius Chlorus. It therefore became necessary to
suppose either that Constantino's artists had done this, which
was hardly tenable, or that a few of the medallions had been
used by one or more emperors of the third century in some
earlier arch from which they would have been once againremoved to the arch of Constantine, thus reuniting them once
more with the rest of the medallions. This hypothesis also
shows into what straits the Constantinian theory was forcingthe best critics.
During this time no serious objection was raised to the
attribution to Trajan and his Dacian victories of the four greatbattle scenes from a colossal frieze, now set into the passage-
way and the ends of the attic.
As for the sculptures of late date and poor style, they had
all been ascribed to Constantino's artists: the Victories, the
River Gods and Seasons of the spandrels; the keystones; the
frieze; the sculptured pedestals of the columns. Quite re-
cently, however, a dissenting voice was raised in regard to the
frieze, the greater portion of which, including the triumphal
procession, is ascribed by Mr. Wace to an arch or some other
monument of Diocletian, a theory which would involve the
wanton destruction of this monument only ten or fifteen yearsafter its constructon. 1
This summary of the present attitude of critics toward the
arch shows that the question has been attacked merely from
the side of the aesthetic qualities of the sculptures, if we
except a few valuable observations by Peterson on the main
cornice and its columns and pilasters. In my own examina-
tion, the question will be studied from different points of view,
and particular stress will be laid on the structural and technical
problems presented both by the sculptures and by the architec-
tural details. The solution which this study suggests will be
!Wace in B.8.R. III, p. 270 ff. Cf. Monaci in .B. Com. Horn. 1900, p. 75
ff. and Atti Pont. Acad. di Arch. 1901, p. 107 ff. and 1904, p. 3 ff.
THE ARCH OF CONSTANTINE 373
tested by the historical, literary, and traditional evidence : onlysuch aesthetic questions will be raised as bear upon the prob-lems of chronology.
In order to clear the horizon, the dedicatory inscription mustfirst be examined. It would seem to state in precise termsthat the arch was built for Constantine, and to make it futile
even to discuss the question, unless we admit that this was oneof the not unknown cases in which a restorer claimed to be thebuilder. But it is not necessary to have recourse to any such
hypothesis. Paradoxical as it may seem, it is out of themouth of the inscription itself that I can prove that the archexisted long before Constantine. It reads (C.LL. VI, 1139):
IMP. CAES. Fl_. CONSTANTINO MAXIMOP. F. AVGVSTO S. P. Q. R.
QVOD INSTINCTV DIVINITATIS MENTIS
MAGNITVDINE CVM EXERCITV SVO
TAM DE TYRANNO QVAM DE OMNI EIVS
FACTIONE VNO TEMPORE IVSTIS
REMPVBLICAM VLTVS EST ARMIS
ARCVM TRIVMPHIS INSIGNEM DICAVIT
Now, if we compare this inscription with others on triumphalarches, of which I give typical instances in a footnote,
1 two
1 The simplest form of arch dedication is that on the arch of Titus : Senatus
Populusque Romanus divo Tito dim Vespasiani f. Vespasiano Augusto. Acontemporary example of the fuller form corresponding roughly to the formula
on the arch of Constantine is that on the destroyed arch of Titus in the Circus
Maximus : Senatus Populusq. Romanus imp. Tito Caesari divi Vespasiani f.
Vespasiano Augusto pontif. max. trib. pot. X, imp. XVII, cos. VIII, p.p., principi
suo, quod praeceptis patris consiliisq. et auspiciis gentem ludaeorum domuit et
urbem Hierusolymam, omnibus ante se ducibus regibus gentibus aut frustra
petitam aut omnino intemptatam, delevit. Both types appear, in the next
generation, on the arches of Trajan. The simpler formula, slightly expanded,is at Beneventum : Imp. Caesari dim Nervaefilio Nervae Traiano optima Aug.Germanico Dacico pontif. max. trib.potest. XVII, imp. VII cos. VIp.p. fortissimo
principi, Senatus P. Q. R. The fuller form appears at Ancona, in which, after
the imperial titles, we read : promdentissimo principi Senatus P. Q. R. quod
accessum Italiae, hoc etiam addito ex pecunia suaportu, tutiorem navigantibus
reddiderit. In the previous period, we find the longer formula represented on
the arch of Claudius in Rome recording the conquest of Britain : Ti. Clau[dio
374 A. L. FROTHINGHAM
differences will be particularly noticeable: that there are no
chronological or triumphal titles given to Constantine, as is
customary especially after the second century, and that the
last line, in which the arch is mentioned, is an addition to the
normal formula, which is unique in Rome and, in fact, in all
Italy. In all other cases the inscription is a mere dedication,
without particularizing what is dedicated. Normally the in-
scription would have ended with the word armis. There must
be some reason for this break with traditional usage, a break
which places this arch in a category of its own, and this reason
must be sought for in the wording of this additional line.
What is the exact meaning of arcum triumphis insignem? The
unprejudiced Latinist would unhesitatingly translate it "this
arch famous for its triumphs." Why has it not been so under-
stood ? Because such a translation would not square with the
supposition that the arch was built for Constantine, since
Constantine had not had even a single triumph, much less
several triumphs. His triumphal entrance into Rome after
the victory over Maxentius was merely a popular ovation,
not a triumph, which is a matter formally voted on and decreed
for certain specific deeds, including the enlargement of Roman
Drusi /.]... Senatus Po[pulusque] Ro[manus q~\uod reges Britanniai XIdevictos sine ulla iactura in deditionem acceperit gentesque barbaras trans
oceanum primus in dicionem populi Eomani redigerit. Among the simpler and
shorter formulas of the Augustan age, the arch at Rimini represents the fuller
form (C.I.L. XI, 365), showing. that the arch commemorated the building and
repairing of Italian highways and ending: celeberrimeis Italiae vieis consilio
\_et sumptib~\us suis muniteis. The style thus inaugurated by Augustus in his
early years, and which, as we have seen, was continued until the close of Trajan's
reign, was not discontinued under the later Antonines, for it appears on the arch
of Septimius Severus in the Forum : Imp. Caes. Lucio Septimio . . . et Imp. Caes.
M. Aurelio L.fil. Antonino . . . ob rem publicam restitutam imperiumque populiEomani propagatum insignibus virtutibus eorum domi forisque, S. P. Q. R.
In all these cases the monument bearing the dedicatory inscription is left
unmentioned.
It is a fact that has some bearing on the present case that the arch of
Augustus at Fano was restored under Constantine and rededicated to him a fewweeks after his death, between May 22 and September 9, 337 : Divo AugustoPio Constantino patri dominorum curante L. Turcio Secundo, etc. The original
dedicatory inscription of Augustus was left when the new dedication was added.
This constitutes the main difference between the arch at Fano and the arch in
Rome, whence the original inscription had disappeared about two hundred andfifteen years before the Constantinian dedication was added.
THE ARCH OF CONSTANTINE 375
territory and the conquest of foreign foes, none of which Con-stantine could claim. Yet, when the arch was dedicated to
him, it was famous, noted, for its connection with several
triumphs. Not even by the greatest stretch of the imagination,or by granting a breach of immutable Roman law and custom,can one regard this expression as referring to Constantine.
On the other hand, it is easy to see how the senate, by takingan arch already built, already used as a triumphal arch, and
rededicating it to Constantine, could by this subterfuge honor
the emperor without breaking the law. 1
The second peculiarity to which I referred is the absence in
the inscription of any chronological and triumphal titles such
as are ordinarily given to emperors on their triumphal arches
under the middle and later empire. This is the more inexplica-
ble because in the latter part of 315, when the arch is supposedto have been dedicated, Constantine had already been given in
inscriptions of 314 and 315 such triumphal titles as Germanicus
Maximus, Gothicus Maximus, Sarmaticus, Britannicus, Persi-
cus, Adiabenicus. 2 In the absence of chronological data in the
dedication itself, the only reason there has been for the selection
of 315 as the date of the arch has been the supplementary
inscriptions in large letters lower down. On the northern face
are: VOTIS X on the left pylon arid VOTIS XX on the
right pylon ; and SIC X SIC XX in the corresponding posi-
tions on the southern face. It has been supposed that these two
expressions were undoubtedly connected with the decennalia of
the emperor, which took place on July 25, 315, and that they
expressed the hope that his twentieth would be as auspicious as
his tenth anniversary. It seems curious that no scholar should
have tested the accuracy of such a conclusion, but that all have
1 See the condemnation of Constantine by Ammianus Marcellinus (XVI, 10)
for breaking this Roman tradition by erecting arches in Gaul to celebrate vic-
tories in wars that were civil or within Roman territory. The proper theory is
referred to in Pliny's Panegyric of Trajan, where Domitian's construction of
arches without corresponding additions to Roman territory is condemned.
2 See Ferrero, in Atti Acad. Sc. di Torino, XXXII, p. 837 ff. Cf. C.I.L.
VIII, 10064; XI, 9
;also Pauly-Wissowa s.v. Constantinus. It is still asserted
that Constantine did not receive the title Maximus, which is given to him on the
arch, until 315, but Cagnat himself (p. 483) acknowledges that Babelon has
proved (Melanges Boissier, p. 53) that he had it as early as October 312.
376 A. L. FROTHINGHAM
followed one another unquestioningly. The slightest inquiryl
would have disclosed the fact that neither in the case of the
VOTIS or in that of the SIC was such a rule actually followed
by Roman custom. In the case of Probus (276-282), thoughhe reigned for only about six years, we find on his coins the
expression VOTIS X et XX FEL. Constantius Chlorus, whowas Augustus for only about a year, has on his coins VOT. XXSIC XXX. Gratian (361-389), at the most liberal allowance,
can be given only 28 years, yet his coins have VOT. XXXMVLT. XXXX. Of emperors whose coins have VOT. XXMVLT. XXX Constans reigned only five years, Valentinian
II about eleven, and Valens about fifteen years. The expressionSIC X SIC XX is used of several whose reign was much under
ten years Galerius, Maximinus, etc. Numerous examples can
be gathered from Cohen, Eckhel, et al. The conclusion is that
the expressions SIC X or VOT. X were used or could be used
of an emperor during any year of his reign from the second to
the tenth.
There is, then, no ground whatever, on the basis of these
expressions, for dating the dedication of the arch of Constantine
in 315. It could have happened just as well in 314 or 313; or
at any time, in fact, after Constantino's victory over Maxentius
in October 312. This brings us back to the question of the
absence of any triumphal titles in the dedication. If in 314
and 315 Constantine had assumed the titles I have enumerated
above, and if they are not given in the dedication, the logical
inference would be that the date of the dedicatory inscription
antedates 314. I would, therefore, suggest the year 313. Aswill appear later, the work actually done on the arch by Con-
stantine's artists was not so extensive as to make it necessaryto allow more than a few months for its execution.
We may conclude then, merely from the dedication, that the
arch, already associated with several triumphs before the time
of Constantine, was dedicated to him in 313.
Now, an arch, in order to be associated with several succes-
sive emperors, would have to be built originally by or dedicated
to an emperor who suffered after death the memoriae damnatio,
1 An examination of the index of Cohen-Babelon would be sufficient to estab-
lish the baselessness of this imaginary chronological certainty.
THE ARCH OF CONSTANTINE 377
which entailed the casting down of his statues and the erasure
or destruction of the inscriptions in his honor. In the case of
such a triumphal arch, the elimination of the dedicatory inscrip-tion would be supplemented by the destruction of bronze quad-
riga, imperial statue, trophies, triumphal frieze, and any other
decorative features that connected the structure very clearlywith the person and career of the emperor. It would then be
a mutilated civic monument unclaimed and undedicated, which
could be adapted to temporary or miscellaneous purposes, and
could at any time be rededicated. To which of the emperorswith both a triumphal record and the stigma of a memoriae dam-
natio can the construction of the arch of Constantine be ascribed ?
This question, which it would seem almost hopeless to ask, is
answered with unexpected clearness by the famous topographical
Colosseum. Arch of Arch of Temple of
Constantine. Titus. Jupiter Stator.
FIGURE 2. PART OF THE HATERII RELIEF, SHOWING THE " ARCH OF CON-
STANTINE" AS IT WAS IN THE TlME OF DOMITIAN.
relief from the tomb of the Haterii (Fig. 2), known to all
Roman scholars as a corner-stone of Forum topography, which
reproduces the principal buildings along the early part of the pro-
cessional route from the funeral ceremony in the Forum to the
mausoleum on the Via Labicana. The first building is the
temple of Jupiter Stator at the head of the Via Sacra;^
the
second is the arch of Titus ; the third is an arch hitherto uniden-
tified; the fourth is the Colosseum. The artist indicates, as
378 A. L. FROTHINGHAM
clearly as possible, that the arch of Titus is in the foreground,
and that the unknown arch, by its smaller size and lower relief,
is in the distance, close to the Colosseum. It is given in profile
and its fagade has free-standing columns. There is a sculptured
frieze encircling the entire arch under the main cornice, and the
attic is crowned by an imperial triumphal quadriga. An imagi-
nary niche or arcade is cut in the end of the arch for a statue of
the Mater Magna, an indication that her temple was in this
section of the Palatine. Every one of these characteristics
suits the arch of Constantine. This unidentified arch stands
about where it does, faces in about the way it does, and has the
same design, in so far as it can be seen from the end. (Compare
Figs. 2 and 3.) One of the unrecognized facts about the arch
of Constantine is that it probably had a sculptured frieze under
its main cornice which was torn away.'
If the arch on the
Haterii relief is not the arch of Constantine, what is it ? Not a
trace of any arch has been found in the excavation of this
immediate neighborhood, nor is there any possibility that it
could have stood anywhere except about where the arch of
Constantine now stands. A photograph taken with the arch of
Titus in the foreground to the right, with the Colosseum in the
middle background, would include between them the arch of
Constantine seen almost in profile.
What is the date of the arch on the Haterii relief ? Therelief has been generally conceded to be Flavian, or, more
specifically, Domitianic. 1 As the arch of Titus is reproduced,which was finished by Domitian, it can hardly be earlier. Asthe relief represents the funeral ceremony and the opening of
the new family mausoleum, and as busts found in the mau-soleum are generally conceded to be of distinctly Flavian art,
it cannot be later than Domitian. Consequently, the unidenti-
fied arch must belong to the reign of Domitian. The use of
free-standing columns at this early date may be objected to.
It has been supposed that only engaged columns were used in
1Helbig, Fiihrer, Nos. 670-675; Crowfoot in J.H.S. 1900; Benndorf-
Schoene, 343-345; Wace, Frag, of Rom. Hist. Rel.' in S.S.B. Ill, 3. Judg-
ment has been based mainly on the technique of the busts found in the
mausoleum, but that of the reliefs is also convincingly Domitianic. It is myopinion that the group of four divinities is somewhat later, possibly Hadri-
anic or Antonine.
THE ARCH OF CONSTANTINE 379
FIGURE 3. EAST END OF THE ARCH OF CONSTANTINE.
(With inserted medallion and frieze cut long after construction.)
380 A. L. FROTHING PIAM
the design of arch fagades until the time of Hadrian. I have,
however, myself called attention to their use in the early part
of Trajan's reign. So far as we can judge they had not been
introduced in any arches under Augustus, Tiberius, Claudius,
or Nero. The arch of Titus, begun before Domitian's accession,
did not have them. But in this arch on the Haterii relief they
appear, and the inference would be that it is to one of the con-
summate artists who worked for Domitian that the innovation
was due that was slowly to revolutionize arch design. More
than this, I may say that several years before I had identified
the arch of Constantine as an arch of Domitian I had concluded
from a study of the coinage of Domitian that free-standing
columns were used in his triumphal arches. As he made him-
self notorious for the number and magnificence of his triumphal
arches, more than any emperor either before or since, it would
have been natural for his architect to innovate in their design.
The equation, then, can be stated as follows: arch of Con-
stantine = unknown Haterii arch = an arch of Domitian.
The evidence of the inscription and of the Haterii relief,
which might seem to be conclusive in themselves, had, how-
ever, nothing to do either with my first doubts or with mygradually acquired certitude as to the pre-Constantinian date
of this arch. This certitude I gained absolutely from the
study of the construction, and it was based entirely on techni-
cal grounds which showed me how impossible the Constantinian
date was. Only after this conviction had been gained and its
details were being carefully worked at and sifted did I see howboth the inscription and the relief fitted in with my revolu-
tionary idea and gave to my structural argument the seal of
historical corroboration. If the objections to the Constantinian
date which I enumerate on p. 368 are now reread, it will be
evident that they can all be explained by my proposed Domi-tianic date.
As a preliminary to the technical study it will be necessaryto describe how a triumphal arch was built. The materials
varied at different times and in different regions, but one rule
always holds good, because it was the orthodox traditional
method handed down from Greece to Rome : that all the
decorative work was done on the monument itself after con-
THE ARCH OF CONSTANTINE 381
struction. We are accustomed in modern times to the habit of
cutting the ornamentation both figured and decorative
before setting it in place. So it must be reiterated and empha-sized that friezes, medallions, rectangular reliefs, keystones,coffered ceiling, cornices, spandrel groups, were all planned,and the blocks or slabs on which they were to be carved werebuilt up with the structure and left rough, with just the proper
projection from the mass, arid were not touched until the con-
struction was completed. Then the decorative work was
begun, at the top : first the carving and then, at times, the
coloring. This preliminary will make it easy to explain the
real relation of the sculptures on the arch of Constantine to
its structure, and to show how untenable is the current hy-
pothesis.
First, however, a few more words as to the structure itself.
The official or central Roman school as distinguished, let us
say, from the Campanian school, or from provincial schools
like those in Northern Africa and Syria began by building
triumphal and memorial arches of solid blocks of travertine
and tufa and then of travertine alone. This was in the pre-
Augustan and Augustan age.1 Before the death of Augustus,
the spread of the use of decorative sculpture on arches made
artists adopt a facing of marble slabs and blocks covering the
travertine, that should allow of the desirable beauty of detail
impossible in the coarser stone. Beginning with a thin veneer
the marble facing became gradually heavier. In the time of
Constantine and for some time previously the core behind the
facing had ceased to be travertine and had become rubble,
concrete, and brick. This is exemplified in the Janus arches of
the Forum Boarium and at Saxa Rubra, near Rome. In the
arch of Constantine we find the earlier technique of the traver-
tine core, and among existing monuments a close analogy is to
the arch of Trajan at Beneventum. This in itself argues a
pre-Constantinian date for the structure. In the parts where
brickwork is added we find Constantinian work on our arch.
1 Pre-Augustan examples are at Spoleto, Aquino, Trieste, Aix-les-Bains,
Carpentras, etc. Augustan examples are at Aosta, Verona, Rimmi. Note the
thin veneer at Aosta (Porta Praetoria), which is paralleled atSpello in the pre-
Augustan gates.
382 A. L. FROTHINGHAM
It is also necessary to note that Roman builders were ex-
tremely particular not to break the course lines of their masonry,
especially in the facing blocks or slabs. The sculptural decora-
tion was not allowed to interfere with this regularity. The
course lines were made to correspond to the top and base lines
of the reliefs. This was easy when, as was nearly always the
case, the marble facing that was left plain was built up togetherwith the projecting facing that was to be worked by the sculp-
tors. In the unusual cases, in later Roman times, when alreadyfinished decorative units taken from earlier monuments were
embodied in the new construction, as is supposed to have been
the case in the arch of Constantine, it would not be difficult to
follow the same rule. No architectural critic would hesitate
to deny that a Roman architect could have preferred in such a
case to zig-zag his course lines rather than take the trouble to
gauge their height by his material.
But we find that, in order to incorporate the sculptured slabs
into the arch of Constantine the architect was obliged in some
cases to cut into the course above for the length of the sculp-
tured slab ;in other cases to substitute a wider block in that
course with an offset in order to have it set down on to the
sculpture ;in still other cases, to supplement this by the addi-
tion of a small cornice strip at the base. The obvious and im-
perative conclusion is that in such cases the sculptures were
inserted in an already existing structure and could not possiblyhave been built up with it. Again, no competent architect
could decide otherwise. We shall examine presently the ex-
amples of each of these methods of insertion.
Before proceeding let me recapitulate the main reasons againstthe Constantinian date.
(1) It does not explain the series of imperial military busts
crowned by Victories set into the masonry of the minor arch-
ways. The presumption is that they represent emperors, and
that they antedate Constantine.
(2) It does not agree with the fact that the majority of the
sculptured decorations were inserted into the structure of the
arch at some time after the construction. Any architect fa-
miliar with Roman work can see this.
(3) It does not explain the terrible damage done to the main
THE ARCH OF CONSTANTINE 383
cornice while the sculptures of the attic were so little damaged,but such damage could easily have been caused in castingdown the groups on the attic and the attic inscription of
Domitian.
(4) It does not explain the use, in the recut sculptures, of
heads of emperors other than Constantine ; a fact explicable
only on the supposition that the arch was connected with these
emperors.
(5) It does not agree with the fact that the triumphal frieze,
which is even earlier than Constantine, is cut in the already
existing masonry and was neither provided for in the designnor brought from another monument.
(6) One is unable to explain, with this theory, how the
spandrel decoration came to be drafted on a preexistingstructure.
(7) It is, we have seen, contradicted even by the dedicatory
inscription.
(8) It is contrary to Roman law and custom.
(9) It is contrary to conclusions based on the Notitia and the
Haterii relief.
We shall now proceed to the technical analysis, beginning;with what is perhaps the simplest problem, that of the -end)
medallions.
The End Medallions (Figs. 4 and 5). In each of the ends
there is a medallion, on a level with the eight medallions of the
two fronts. They are in a later style and were evidently an
imitation of the series of eight. It has been assumed that theyare of Constantinian workmanship, though it has been grudg-
ingly granted that their art has pre-Constantinian elements. :
As a matter of fact it seems like defying the elementary stand-
ards of criticism to assert that they belong to the same time
and school as the frieze or the spandrels. They seem hardly
later than the middle of the third century, and might belongto the time of Severus Alexander. A comparison of these
horses with those in the triumphal frieze and the Siege of
Verona will illustrate the technical differences. It will be
clear, later on, that even this frieze is pre-Constantinian.
If we examine the relation of these end medallions to the
masonry, it is evident that they were inserted and were not
384 A. L. FEOTHINGHAM
part of the original structure or facing. In order to insert the
Rising Sun medallion at the east end, which was to be cut in a
slab too' short to correspond fully to four courses of the facing,
the architect first inserted at the bottom of the cut which he
made a narrow cornice strip, to serve as a decorative base.
FIGURE 4. EAST END MEDALLION," SOL INVICTUS."
This is an evident insertion because it was against Romancustom to carve such mouldings in separate blocks. Theywere cut either in the top or in the bottom of a wide course,
as can be seen without leaving this arch, for instance, in the
moulding below the frieze on this same east end. But even
with this inserted strip the slab was found not to reach to the
level of the fourth course. The architect, therefore, seems to
THE ARCH OF CONSTANT1NE 385
have removed the facing above it; not only two blocks of the
next course but the central epistyle block. He then shortened
one of the blocks in order to admit of a new block that should
project on both sides of the new medallion. In being put back
the shortened block was injured. A new block was cut so as
to fit down on to the medallion, and the change in the course
line was almost hidden by the thin porphyry framework now
disappeared which was brought up about to the regular course
FIGURE 5. WEST END MEDALLION, "DIANA."
level. The jags cut in the block were plainly visible, however,
at either end. After this the epistyle block was put back with
some slight abrasions. On the right end the upper and lower
facing blocks were not cut, but the medallion slab was cut away
to fit them and the irregularity was concealed by the porphyry
facing of the frame.
In the Moon medallion on the west end, the process was
reversed. The two slabs that compose it (each of the other
medallions is on a single slab) were longer than was needed,
and instead of cutting them down to suit the coursing of the
facing slabs, the two slabs of the course above were cut into.
The base-moulding also was not separate, as on the east end,
386 A. L. FROTHINGHAM
but was cut in the slabs of the medallion. The numerous ir-
regularities seem to show that the insertion was done quite
late and led to considerable disturbance of the entire facing.
Also, when the surface was cut down to form a square frame
filled with some richly colored marble, a queer effect was pro-
duced by the narrow rim of the slab left on the right side
against the courses.
The conclusion is, on technical grounds, that these medallions
were inserted, in the rough block, into the structure of the pre-
existing arch and then carved in imitation of the other medal-
lions ; on stylistic grounds this happened before the time of
Constantino.A. L. FROTHINGHAM.
PRINCETON, NEW JERSEY,
MAY, 1912.
{To be continued.)
%rubatalogical Inotitute of amtdrca
WHO BUILT THE ARCH OF CONSTANTINE?
II
THE FRIEZE
BEFORE continuing the detailed examination of the sculp- tured decoration in its relation to the structure of the arch, which was barely begun in my first paper (A.J.A. XVI, 1912, 3, pp. 368 ff.), I shall give a brief report of the work done on the arch during the intervening year. In October, 1912, I presented to the International Archaeological Congress in Rome
my theory that the arch was built by Domitian, mutilated at his death, restored and used as a general arch of triumph until re- dedicated to Constantine. Thanks to the cordial and efficient
cobperation of Comm. Corrado Ricci, it was possible to pre- pare in the course of only four days special casts and photo- graphs for the Congress. But as they illustrated only the lower
parts it seemed afterwards indispensable to test my theories by a close examination of every foot of the surface up to the top of the attic by means of scaffolds. Only by this searching study could the many puzzles be solved. Comm. Ricci granted his permission, and the Office for the Preservation of Monuments
placed its corps of expert scaffold builders at my disposal for several weeks in April and May, 1913. A several-storied mov- able tower on wheels made it possible to return again and again to the same detail. Above it permanent platforms were built on the attic. No part of the surface was left unexamined or untouched. Even the thickness of the slabs was ascertained wherever there were cracks or holes allowing the insertion of a slender steel rod. Casts and photographs were taken of details that bore upon the problems under discussion. An
experienced archaeological architect, G. Malgherini, took the measurements and profiles for a detailed series of architectural American Journal of Archaeology, Second Series. Journal of the 487
Archaeological Institute of America. Vol. X VII (1913), No. 4.
488 A. L. FROTHINGHAM
drawings; finding, incidentally, that what had thus far been
published was most inaccurate. An expert examination was
begun of the marble used in various parts, in order to ascertain
any differences in quarry and quality. Archaeologists and art critics were freely invited to take part in the examination and to test my theories and statements.
The arch has, therefore, been for the first time studied with the care and closeness that it deserves. It was inevitable that new problems should arise in the course of this study, and yet, while I found occasion to modify some of my opinions and to make some reservations,1 it has not brought any fundamental
change in my views. I have gathered a greater wealth of de- tail to prove the pre-Constantinian origin, and even the pre- Constantinian date (third century) of that latest addition to the structure, the attic. The materials have been gathered for a special volume; what will be used in these articles is only the part that bears on the problems of structure and age.
In regard to the contention in my previous article (p. 375), that it would have been against Roman law and custom to have built an arch to Constantine for victories in a civil war, the force of my argument was impaired by a printer's error which made me say that Ammianus Marcellinus condemned Constantine for erecting arches in Gaul to commemorate in- ternecine victories. What I actually wrote, of course, was not Constantine, but Constantius, his son, and if the contemporary historian condemned Constantius for such an improper inno- vation, it is unthinkable that he should have done so with a new arch of Constantine of this very character before his eyes.
In my previous article I examined only the two late medal- lions on the east and west ends. I shall now examine the frieze.
The frieze is not continuous, as in the arch at Beneventum, but is in sectionis, as in the arch of Septimius Severus. Though it extends around the entire arch it has no unity either in
height or style. There are six sections: one over each of the
1 I found no travertine core (p. 381); everything within sight or touch is marble of various qualities up to the attic, where the structure is later and changes to brick and concrete. This does not, however, exclude the possibility of the use of travertine in the mass of the masonry, but it cannot be proved.
THE ARCH OF CONSTANTINE. II 489
minor arcades on each face and one across each end. Those on the east and west ends belong to a triumphal procession, with an emperor entering Rome, and they overlap around the four corners on to the main faces as far as the first pilasters. The two sections on the south face represent victories leading up to the triumph; those on the north face give the popular festivities in Rome after the triumphal entrance. The univer-
sally accepted theory has been that all the parts of the frieze date from Constantine's time. The two scenes on the south are interpreted as the Capture of Verona (or of Susa), on his
way to Rome, and as the defeat of Maxentius at the Milvian
bridge. The north scenes are then his proclamation to the
people from the Rostra in the Roman Forum and his distribu- tion of the Congiarium after his triumphal entrance.
A dissenting voice has been raised by Mr. Wace 1 in a very interesting paper. Largely on the basis of the substitution of a second imperial head for the first one in three out of the six reliefs and also of the fact that Constantine did not celebrate a triumph, he ascribes the four sections of the triumphal pro- cession, the Rostra proclamation and the Congiarium, to the next previous emperor who enjoyed a triumph; that is, to Dio- cletian. He leaves to Constantine only the two south scenes: the capture of Verona and the Battle of the Milvian bridge, because here he finds that there is no substitution of a second imperial head. He thinks that the four Diocletianic reliefs were removed from some triumphal monument of that emperor and finds substantiation of this in the breaks that run con-
tinuously along the base line of the reliefs, detaching them from the structure of the arch. I do not know how much attention has been paid to Mr. Wace's theory. I found my- self obliged to discard it as more than improbable. The same break which runs along the base line of the reliefs he thinks Diocletianic runs also along the base line of those he thinks Constantinian. In the Battle of the Milvian bridge, where he does not believe the emperor to have been present at all, there are not only remains of the emperor's figure, but proofs of the substitution of a second imperial head; which, according to
1 A. J. B. Wace, ' Studies in Roman Historical Reliefs' in the Papers Brit. School at Rome, IV, p. 270 sqq.
490 A. L. FROTHINGHAM
Mr. Wace's own theory, would be fatal to an ascription of this scene to Constantine. This one fact is sufficient to make his
theory untenable. It has also against it facts of style and
technique. He grouped the Congiarium and Rostra scenes, which are of one style, with the Triumphal Processions which are of quite different technique; whereas the two battle scenes, which he places by themselves, are undoubtedly by the same hands as the Triumph! Nor does it seem at all probable that Constantine should have allowed the mutilation of a monument built only ten or twelve years before in honor of Diocletian.
We may then conclude: (1) that no part of the frieze came from a monument of Diocletian, and (2) that, since both the basal break and the mutilation of the emperor's head are char- acteristic of all the reliefs on the main fronts, they must all have been treated in this way at the same time.
What must be asked is this: Is the mutilation of the em-
peror's heads sufficient to exclude the Constantinian origin of all these frieze reliefs ? Does the break along the base line show that they were brought from some other monument; or does it show, on the contrary, that they were carved in situ and that the decorative work underneath the break was added? Does not the evident stylistic difference between the Congia- rium and Rostra scenes and the rest show that these two scenes
belong to a different period and to another emperor than the
triumphal and battle scenes ? If all the scenes are pre-Con- stantinian and attributable to more than one emperor, does their technique afford any indication of their date ?
These questions will be answered in the course of an exam- ination of each relief, beginning with the scenes preceding the
triumph. I. SOUTH FACE. (a) Siege of Verona (Fig. 1). The
effect of thebase-line cut just under the feet and cutting away the base on which they rested, though far riskier, was but little more damaging than in the Battle of the Milvian bridge; which is remarkable, seeing that there were so many feet almost or entirely in the round. Only one figure lost its feet: the soldier advancing alone to the right under the city walls. The reason for his mutilation would seem to be that the block
immediately beneath him took an upward curve instead of
THE ARCH OF CONSTANTINE. II 491
following in a straight line par- allel with the left-hand and central cornice blocks. This in- volved a miscalculation, and the
necessity of cutting away the feet of the only figure in the lower
part of this block in order to secure, according to which theory one adopts, the insertion either of the frieze block or that of the
spandrel block. It will be noted that an attempt was made to conceal the shortening of the
legs by thinning them off in a
very clumsy way. There are two details which
bear on the question as to whether this section was inserted or was carved in situ: these details are the bow of the middle archer and the lance of the soldier below him. This middle archer was carved on two slabs of the upper course. His left arm is extended so that it is mostly on the right- hand slab, holding the bow, of which the top remains against the frame. The right arm, hold-
ing arrow and string, with its forearm in the round, was, with bow and arrow, partly on one and partly on the other slab. It would have been impossible to move these slabs after carving. The damage to these delicate details which we see at present was presumably done not in ancient but in mediaeval times. Still, as it might be argued that
i3
r;? AB
z
C
Q Fr(
C
z
z C
J
C: Q
492 A. L. FROTHINGHAM
it could have been done by transportation, it is lucky that the soldier below the archer, who is carved both on the upper and lower courses, is holding a lance whose delicate shaft is cut in the round against the upper slab and connects closely and perfectly with the continuation of the shaft on the lower slab. Here at least the perfect preservation of this detail in unrestored condition is enough to prove that the slabs were all carved in situ, and that the damage to the archer's bow was not due to transportation. The head of the emperor, who stands to the left, with a flying Victory behind him, has not been treated as in the other cases to be studied. It is mutilated, but there was no provision for a substitution in ancient times of a second head. It is extremely difficult to decide what this difference means. Does it mean that this section does not belong to the same time and refer to the same
emperor as the east and west sections ? If the head was that of Constantine, must we not infer that the mutilation was acci- dental? But how can the accident theory seem plausible in the face of the fact that every other head in the composition is intact? It is easier to suppose that the head was made
unrecognizable and that it was by mere carelessness that no second head was substituted. This supposition will be con- firmed by the examination of the emperor's figure in the cor-
responding scene of the Victory of the Milvian bridge, as the two scenes certainly belong to the same series.
Of course this theory would exclude the title " Siege of Verona " as the subject of the relief.
SOUTH FACE. (b) Battle of the Milvian Bridge (Fig. 2). This scene is more seriously damaged than any of the series, especially where the emperor stood. The photograph reproduced in Figure 3 was taken for the purpose of showing the line of cleavage along the base, which is continuous, and just as clearly as in the other cases posterior to the carving. The peculiar thing about the emperor in this scene is that he has totally disappeared and that "his place knows him no more." He stood between the Virtus who advances to the right over the arch and the river god, and the winged Victory beyond, whose head is turned backward as she leads the emperor onward. Of the emperor himself who occupied the vacant
THE ARCH OF CONSTANTINE. IT 493
space between Virtus and Vic-
tory, there remain only the two feet and the outline of some
drapery against the background below the place where the knees were. Also, the fractures show where the outlines of the figure were. Especially conclusive is the cavity made where the head was, showing the intention, at least, of substituting a second head. It is curious that nobody appears to have noticed these facts. They either think, as Mr. Wace does, that the em-
peror was not present or else
they see him in one of the soldiers in front of Victory. That the emperor was actually represented and that his head was hammered away as in the
Triumph, the Congiarium, and the Rostra scenes, completes the upsetting of Mr. Wace's
hypothesis. In both of these scenes, while
the figures are heavy and
clumsy, there is considerable action, energy, and some expres- sion. This is particularly true of the archers, of some of the horsemen, and of the Virtus. The marking of the pupils of the eyes is one of the technical points of difference between these reliefs and those of the north face, where the eyeballs are smooth, and there is no
attempt, as here, at expression.
OF
CO.
lb
A4.
z
fQ
Z
0
H
5?
z
H
z
Fr•
0 0
0
0 0T
•5
494 A. L. FROTHINGHIAM
A good argument against the supposition that these scenes
might have been brought from another monument is that they are so composed as not to have fitted any position but one of
exactly their present length. One can imagine all the other four sections of the frieze as longer or shorter without funda- mental disturbance, but the Milvian bridge scene in particular cannot be imagined as changed by addition or subtraction. It was designed for its present place, without a doubt. If it is
pre-Constantinian, then the arch is also pre-Constantinian. I
. -A?A4
41
'Io-Opp
FIGURE 3.- DETAIL OF FIGURE 2.
keep on using the term " Battle of the Milvian Bridge," though if the scene is pre-Constantinian, some other battle at a river must be selected as the theme.
II. EAST FRIEZE. Triumphal Entrance of the Emperor, pre- ceded by the Army (Fig. 4). No continuous cornice frames the top; there is only a cornice strip under the medallion and connected with it. The cornice that serves as a base to the frieze is an important factor. Its lines, where they are pre- served, are straight; its curved surface is even and of good outline. This is in strong contrast to the corresponding cor- nices on the main faces, and shows an earlier date. The cornice is cut in the same blocks that form the lower half of the figures
THE ARCH OF CONSTANTINE. II 495
of the frieze. It will be noticed that this argues in favor of the
theory that this eastern section of the frieze was carved in situ. Two other arguments for the same theory are: that the frieze is cut in two normal courses of marble blocks, of the same average height as the rest of the building blocks of the arch; and that the figures are cut back instead of projecting in relief from the structural line. This latter argument I have
already used as tending also to show that the frieze was not
planned when the arch was built, as otherwise the blocks on which it was to be carved would have been set in projection, according to the common custom.
The next point is the head of the emperor, seated in the chariot to the extreme left. The head has disappeared; not only that, but the hollow in the neck shows that the original head had in ancient times been replaced by a second imperial head. Now a technical comparison shows abso- lute identity of style between this section of the frieze and the already examined two sections on the south face - the Capture of Verona and the Battle of the Milvian bridge. It would follow from this clear substitution of a second imperial head that the original work could not be Constantinian, and if this is true of the eastern frieze with
_::i??
a i~?B
?i
,1,
0; z S;
S
SI H S S
S
S
S z
Q
S4 0 S
S 0
496 A. L. FROTHINGHAM
Ilk de
Art Iff
00
if 4-.
f
IN
?A 414'
z
.
z
z a
Q 0 o
S 0
12
the emperor's triumphal entrance, then it is also true of the scenes on the south face and the west end. These scenes, therefore, cannot be incidents in the life of Constantine, but must be incidents in the life of whatever preceding emperor was commemorated in this east frieze.
Perhaps the details will give some hint as to who this emperor was: the presence of camels and mules, the type of captured or carried
standards, the costumes. III. WEST FRIEZE. Triumphal
Procession with Captives and Booty (Fig. 5). This face is in not nearly as good condition as the east face, apparently on account of the in- ternal staircase at this end, which seems to have led both in ancient times and in the late Renaissance to considerable reconstruction, re-
modelling, and mutilation. This is shown in several ways: by the modern doorways, the irregular joints, the difference in the size,
quality, and finish of certain blocks, such as those above the frieze on the right; the rough openings cut to give air and light to the staircase. This remodelling involved the in- sertion of antique fragments, espe- cially a cornice block near the base of the staircase, which has been used as an argument for the Con- stantinian date of the arch, whereas it merely shows that the staircase was either built or remodelled at the time of the construction of the
THE ARCH OF CONSTAN TINE. II 497
attic, which was either under Constantine or in the latter half of the third century, if my theory is correct.
The ancient remodelling would seem to have been done when the frieze was carved, and in this case would antedate Constan- tine. A strong indication is in the arrangement of the base, which is entirely different from that of the east frieze. The base cornice is carved on the top of the course below the frieze instead of in the lower course of the frieze itself. The narrow band along the top of the frieze, which existed in the east frieze but was started so far back as to be inconspicuous, is here kept forward. The reason for this, however, is evident. It is that on the east end the figures are smaller and in two rows, superposed in rough perspective, so that the listel must recede in order not to overshadow the upper rear row of fig- ures. This makes it the more noticeable that on the west frieze, where there is a single line of larger figures, the relief is much lower and the listel is flush with the face of the arch, that at the right end there is no listel at all. This is due to the fact that the two blocks in the course above this part of the frieze, which I have already noted as of a different finish (and which are probably of a finer-grained marble), are laid on a level below the rest of the course by four centimetres, just the width of this narrow band, so that in order to keep the figures of uni- form size at this end it was necessary to carry them up to the full height of the slab.
At this point comes a significant observation. Two of the Roman legionaries in this section are carrying standards cap- tured from the enemy. One is surmounted by a nude male
figure; the other by a draped female figure. We might call them, for convenience, Heracles and Nike, as their type is classic. Now the heads of these delicate statuettes touch the lower
edge of the structural block above the frieze. This bears conclu-
sively on the question whether this frieze was carved in situ or was brought from some other monument and inserted in the
masonry. Any architect will, I think, agree that if this block of the frieze had been already carved, it would have been im-
possible to use it in building the arch and to superimpose the block above it without fracturing these figurines. The same
applies in a lesser degree to the crests of the four soldiers' hel-
498 A. L. FROTHINGIIAM
y,
I, ::II
,?
o z S
S
Q
z
0 as
mets. This would dispose of Mr. Wace's theory that the frieze was brought here from a monument of Diocletian.
Against this theory as applied to the other frieze at the east end is the fact that the frieze is there cut in the same block as the base cornice which is an integral part of the struc- ture and can hardly have been
brought from elsewhere. IV. NORTH FACE. (a) Tie
Emperor on the Rostra, ad-
dressing the people (Fig. 6). The break along the base line, instead of being below the feet of the figures, as it is on the south face, cuts across
their legs at the ankles or above them, both here and in the Congiarium scene. In the centre the line of the cut is on a higher level through the base of the Rostra, along the entire length of the cen- tral block of the cornice below. Why is this so? The most plausible explanation seems to be that this was done in order to allow of more width for this cornice block which otherwise would have been so slender as to be
easily broken in the centre. This explanation carries with it the assumption that this cornice block was inserted
after the carving of the frieze,
THE ARCH OF CONSTANTINE. 1 499
and if this block, then the rest of the cornice and the spandrel sculptures. In support of this is the evident fact that the surface of the Rostra is carefully finished except only the
part of it represented on this cornice block, where the surface is quite rough and done at a different and later time. The same crudeness appears in the handling of the upper part of the other two cornice blocks on which the figures of the frieze stand. The feet of the figures, below the break, are not the original feet; that is so clear as not to require argument. Not a single foot is carefully finished, in the style of the rest of the figure. In some cases, especially on the extreme right, there is not even an attempt made to fashion
any feet below the break. In fact the fracture in some cases is so far above the base of the ankle that to carve new feet would have meant attaching them almost to the base of the calf of the leg. About 15 per cent of the lower leg was cut away. In so far as the frieze itself is concerned, and its connection with the cornice, the balance of evidence is in favor of its
carving in situ and its mutilation when the spandrel decoration was added. This question will be more fully studied in con- nection with the spandrels.
The head of the emperor, who stands in the centre, was hammered off, and both background and neck hollowed out to receive a second head, now lost, if we assume that it was ever actually put in place.
NORTH FACE. (b) The Imperial Congiarium (Fig. 7). The emperor, seated on a throne and surrounded by his court, is acclaimed by the populace, while the routine work of handing out the gifts is carried on in four offices on either side. The break runs across the base in a perfectly straight line. It is even more evident here than in the preceding scene that the lower part of the figures has been cut away. I had a special photograph (Fig. 8) taken with a slight downward tilt in order to show that all the figures lost their ankles or even more, and that only in a few cases was any attempt made to replace the lost feet, even in the crudest way. The injury to the legs in some cases included a break higher than necessary. The tech- nique is the same as in the previous relief. There is no ex- pression; the attempt at action is like the attitudinizing of
500 A. L. FROTHINGHAM
hm
II,~ ~ ANt- ' 'mom
_..P
0 Q
Q
0
Q
Q
puppets in a punch and Judy show. The treatment of hair and eyes differs entirely from that of the south face. The art is far poorer; one would be
tempted, stylistically, to call it
post-Constantinian. The head of the emperor has
disappeared. As in the previous relief there have been two suc- cessive heads, the head last lost
having been set in the cavities of neck and background which had been hammered out to a considerable depth.
The same relation to the span- drels and cornice obtains here as in the previous case.
Just a word about an insignifi- cant part of the frieze, which seems always to have been passed over without remark, though it has an interesting bearing on the whole question: I mean the returns of the triumphal proces- sion around the four corners of the main faces up to the pilasters. There is but little room, only for a horseman, a horseman and a
footman, or a couple of foot
soldiers; yet these short reliefs
may help toward certain con- clusions. They are cut in the same marble blocks as the begin- ning and end figures of the main line of triumphal procession, so
they favor the theory that the
processional scenes were carved in situ. Their base line is on the
THE ARCH OF CONSTANTINE. II 501
same level as that of the big frieze of the north and south faces, but their upper line is on a considerably higher level, which, if continued along these faces, would have intersected the base of the medallion frames.
There is another suggestion. Had the designer of the arch
planned to carve a frieze at this point when the arch was built, he would certainly have made the two courses in which it was to be carved of exactly the same height, so that the figures would correspond in size at all points. As we see, this was not done. The top of the second course over the minor arcades is
?? A " i ? . AP All, io, 4111,
AW5 IANIS
EAS?.A tkwor;.
mul
4 ftbo FIGURE 8.-DETAIL OF FIGURE 7.
on a lower level than that of the corresponding course beyond the framing pilasters near the corners, so that any one running his eye along the main face, can see at a glance that the cavalry and infantry in the returns at the corners, and belonging to the
Triumph, reach a higher level. This is another argument against the theory that the frieze was cut when the arch was built, to supplement the projection argument, and favors the
pre-Constantinian date for the structure. If one takes the trouble to analyze the grouping of the three
decorative elements above each minor arcade, - spandrels, frieze, and medallions, --what is the impression one receives? Is it not that the designer was hampered and cramped in some
way ? If he could have set the medallions at any point in the
502 A. L. FROTHINGHAM
structure, why did he set them so low that he was obliged to flatten the circle of their base most ungracefully in order to make room for the frieze? And if he was free to set the frieze where he chose, why did he crowd it down on the arch-
way so closely as not to give himself the room to run the cornice on an even line back of the keystone ? The arch of Severus shows the normal distance between archivolts and cornice. Once it is admitted that the medallions had been in place long before the carving of the frieze and spandrels, the puzzle explains itself.
What seem to be the results of this study of the six sections of frieze ?
In five sections the emperor is present. In each case his head is mutilated; in one of these cases it was made unrecog- nizable and in the other four another imperial head was substi- tuted. As the only remaining section is part of the triumphal procession and is in the same style, it cannot be separated from the rest. Therefore, whatever this mutilation of the emperor entails is entailed for all six sections.
What is this consequence? Unless we suppose the arch to have been mutilated in favor of some emperor subsequent to Constantine, which seems unlikely,' the consequence is the following dilemma. Either the entire frieze was brought from other monuments, dedicated to other emperors, whose heads
1 Did Constantius mutilate his father's arch ? For the sake of argument, and in order to exhaust every possible hypothesis,
I would offer the suggestion, merely as a suggestion to be taken into considera- tion, that when the Emperor Constantius entered Rome in triumph in April, 357, he may have been guilty of mutilating his father's arch, and in some cases have substituted his own portrait for that of his father. Ammianus Marcel- linus describes quite cynically this absurd travesty of the older triumphs, and possibly the emperor distributed a Congiarium and assembled the people in the Forum for a remission of taxes, after having entered in triumphal proces- sion with imaginary captives and counterfeit booty. If it should be admitted that in order to commemorate this event certain scenes on the arch were made to apply to Constantius, then it would still be possible to connect Constantine with the " Siege of Verona " and the " Battle of the Milvian bridge " because the mutilated emperor in these scenes might then be Constantine.
Constantius remained only a month in Rome, however, leaving in May, and in view of his absolute indifference to the city both before and after this visit, and his unpopularity there, it seems hardly worth while to consider this sugges- tion at all seriously.
THE ARCH OF CONSTANTINE. II 503
were removed or mutilated so as to use Constantine's head in their place; or else the frieze was carved in situ and the same
process was employed. The examination has shown that of these two alternatives
the latter should be adopted because of proofs given in more than one relief that transportation of the finished reliefs with- out fracturing details that are still intact would have been
impossible, and these proofs are supported by the improbability of securing from the spoils of other monuments appropriate scenes that would exactly fit the length of the places to be decorated, as well as by other minor arguments that need not be repeated.
If the entire frieze, then, was carved in situ, it was carved in pre-Constantinian times, as is proved by the mutilation and substitution of the imperial heads. Was it all carved at the same time ? Were the original imperial heads all of the same
predecessor of Constantine ? We have found that a well- defined difference in style and technique separates the frieze into two groups. In one group are the battle-scenes on the south face, the two short ends, and the returns around the ends with the triumphal procession. In the second group are the Rostra and Congiarium scenes on the north face. The two groups, having been carved at quite different times, pre- sumably refer to two different emperors, both of whose por- traits were destroyed to be replaced by heads of Constantine.
It would be idle to speculate at present as to the identity of these two predecessors of Constantine, except that on stylistic grounds they can hardly antedate the middle of the third cen- tury. The important and fundamental fact is that the frieze, if these premises are correct, proves the existence of the arch before Constantine.
A. L. FROTHINGHAM. PRINCETON, September, 1913.
Archaeological institute of America
AMERICAN
JOURNAL OF ARCHAEOLOGY
THE JOURNAL OF THE ARCHAEOLOGICAL
INSTITUTE OF AMERICA
VOLUME XIX
1915
CONCORD, N. H.
PUBLISHED FOR THE INSTITUTE BY
Wbt !\umforb $re^NEW YORK: THE MACMILLAN COMPANY
LONDON: MACMILLAN & CO., LTD.
Archaeological
Institute
of America
WHO BUILT THE ARCH OF CONSTANTINE ? Ill
THE ATTic 1
IN the field of Roman historical reliefs the only rival to the
series of eight colossal panels in the attic of the Arch of Cohstan-
tine is the decoration of the arch of Trajan at Beneventum. I
mean, of course, these eight supplemented by the three reliefs in
the Museo dei Conservatori, making a group of eleven a twelfth
being missing and all hfeing supposed to have originally formed
a group of twelve which decorated a monument of Marcus Aure-
lius.2 Mr. Stuart Jones thought that this monument was the
triumphal arch built in this emperor's honor in 176 on the Capi-
toline, for the double triumph over the Germans and Sarmatians.
The twelve reliefs are supposed by him to depict the main epi-
sodes of this double war, and to have been arranged in groupsof four on each main face of the attic and two on each end. I
would entitle the eight panels on the attic as follows :
North Face 1. Adventus Augusti 2. Profectio Augusti3
(beginning at the left) 3. Congiarium P. R. 4. Captives before the Emperor
South Face 1. Rex . . datus* 2. Captives before the Emperor3. Adlocutio 4. Lustralio.
1 For previous papers see A.J.A. XVI, 1912, pp. 368 ff. and XVII, 1913,
pp. 487 ff.
2Strong, Roman Sculpture, pp. 291 ff. and 392 ff.; Stuart Jones, in Papers
Brit. Sch. at Rome, III; pp. 251 ff.; Petersen in Rom. Mitt. 1890, pp. 73 ff.
3 The latest explanation, that this scene relates to the emperor's triumphalentrance into Rome, does not seem to me to suit the treatment of the scene,
which shows the emperor and his suite about to mount horse and take to the
road which welcomes him; a scene frequent on the coins.
4 There are coin types which favor both interpretations: that now commonlyadopted, that the emperor is here dismissing the praetorian veterans at the
end of the war, and that which seems preferable to me, that the emperor is
here assigning a king to some barbarian nation, as was usually done in these
wars. The type of men here portrayed does not seem to me Roman but
oriental. The praetorian interpretation seems to me excluded not only on
American Journal of Archaeology, Second Series. Journal of the 1Archaeological Institute of America, Vol. XIX (1915) No. 1.
2 A. L. FROTHINGHAM
In April-May, 1913, I was able to study the panels of the attic
more closely than any archaeologist had been privileged to do, on
the scaffolding built for me on the attic, as I have described in a
previous article. During the course of more than a week I
handled and examined every detail, and made photographs. I
also examined the interior brickwork, concrete, and stonework of
the attic. My conclusions did not agree with the theory of an
FIGURE 1. RELIEFS AT EAST END OF SOUTH FACE, ARCH OF CONSTANTINE
(a) Adlocutio (wide frame)Lucius Verus series
(b) Lustratio (narrow frame)Marcus Aurelius series
original single arch with twelve attic reliefs, from which these
eight were taken. I tried not to allow my feeling that such an
overloaded attic was inherently improbable to influence myjudgment. It was for internal reasons only that I felt obliged
this account, but because a careful examination of the coins shows that where
the figure at the foot of the tribunal is prominent and with his back squarely
turned to the emperor, and the emperor's hand is extended over his shoulder,
as in the relief, the scene is invariably the presentation of a king. On the
other hand, where the subject is the dismissal of the praetorians, the officer
does not back squarely to the emperor; is in the background, often in smaller
size, and the figures addressed are in uniform, holding standards.
THE ARCH OF CONSTANTINE III 3
to adopt the theory that these reliefs originally belonged to at
least two distinct arches : one an arch erected not to Marcus
Aurelius alone, but to Lucius Verus and Marcus Aurelius, for the
Parthian triumph of 166 A.D.;and the other an arch built ten years
later, in 176, to Marcus Aurelius alone. These reasons are as
follows :
(1) If they had been all prepared for one attic, the heavy mould-
ing which forms the frame for each one and is cut in the same
immense slab, would be of uniform outline and size. This is,
however, not the case, but the frames vary enormously, and do
so not carelessly but so that they fall into two distinct groups,
which can readily be distinguished even at a distance. This can
be seen in Figure 1, which gives the two reliefs at the east end of
the north face. The narrow frame is used in the right panel, the
**
RELIEF
FIGURE 2. PROFILES OF Two TYPES OF FRAMES OF PANELS
wide frame in the left panel. Here the upper part is much re-
stored, but follows the original lines. It will be seen later that
the division of the reliefs into two groups according to the frame-
outline corresponds to that on the basis of style. In Figure 2
I give the profiles of the two types. The difference is too great
to be accidental. It would alone seem sufficient to forbid deriv-
ing both groups from the same attic.
(2) There is a distinct difference in style. The two left-hand
reliefs on the north face show an exquisite finish, a Hellenic ideal-
ism and a grace of attitude and movement, which are less evident
in the two corresponding reliefs on the right side of the same face.
The figures of Roma and Fortuna, from the Adventus (Hellenic)
are given in Figure 3. The two reliefs on the extreme left and
extreme right of the south face are stylistically similar to the left-
hand group of the north face. This is proved, for example, in
the relief on the right end by the charming Hellenic Camillus
4 A. L. FROTHINGHAM
(See Fig. 1, b. and Fig. 6). On the other hand the two other
reliefs on the south face those near the centre are in the same
more typically Roman manner of the right-hand pair on the
north. A glance at Figure 1 will show another difference be-
tween the two series : the base of the left-hand relief is five centi-
metres lower than that of the right-hand relief. Another stylistic
difference which is very marked is the treatment of beards and
hair. That of the "Hellenic" series is shown in Figure 4, represent-
ing the Senate, in the Adventus scene. The figure is a fine exampleof the highly finished style. The beard and hair are in continu-
FIGURE 3. ROMA AND FORTUNA
(Hellenic style)
FIGURE 4. SENATUS
(Hellenic style)
ous sweeping locks, with only a moderate use of the staccato
effects of the deep drill. The very different treatment of the
"Roman" series can be studied in Figures 5 and 9 where the
treatment is coarse, with universally deep drill work and stronger
contrasts.
If we examine the framing of these groups it appears that the
four" Hellenic" reliefs, if I may so refer to them, have the nar-
row frame, while the four" Roman" reliefs have the wide frame.
This again can hardly be a coincidence.
(3) The third point will, I think, make it possible to date these
two series as well as to confirm them. It has to do with the
THE ARCH OF CONSTANTINE III
military standards. In the extreme right-hand relief on the north
face, where a barbarian chief and boy appear as suppliants before
the emperor, the main, central standard has the medallion por-
traits of two emperors, surmounted by Victory (Fig. 5). This
relief belongs to the" Roman" series. On the other hand, the
extreme right-hand relief on the south face has a standard with a
FIGURE 6. HEADS AND STANDARDS
IN SOUTH 4 (Hellenic Style)
FIGURE 5. HEADS AND STANDARDSIN NORTH 4 (Roman style)
single imperial portrait medallion. This is one of the" Hellenic"
series (Fig. 6).
Now Lucius Verus shared the empire with Marcus Aurelius
(161-180) until his death in 169. Marcus Aurelius was sole
emperor from 169 till 177, when he made his own son Commodus
co-Augustus. During this period two triumphs were celebrated.
The first, in 166, was for the Parthian war (161-165) and was in
honor of both emperors. To an arch commemorating this tri-
umph the reliefs of the series to which the standard with the
6 A. L. FROTHINGHAM
double portrait belongs should be referred. They cannot possi-
bly refer to the Germanic-Marcomannic wars when Marcus Aure-
lius ruled alone. The second triumph was in 176, seven yearsafter the death of Verus and one year before Commodus was made
Augustus. Only the series to which the standard with the single
portrait belongs can be connected with this triumph; for MarcusAurelius did not assume the title Germanicus till 172, and that
of Sarmaticus not till 176.
The importance of these images of the emperors on the stand-
ards can hardly be doubted. It was to them that the soldiers
swore allegiance. The well-known passages in Tacitus 1 indicate
that these portraits were movable and could be exchanged in
the medallions on the standards on the accession of a new em-
peror. The presence of one or of two portraits on a standard
may be taken as absolute proof that at that time the empire wasruled by a corresponding number of emperors. So far as I know,this deduction has never been made, nor the importance of these
images appreciated in their historic bearing. I have other cases
where they give equally important results in the way of historic
identification.
Two questions may here be asked. The first is: Why do not
both emperors appear in the first series, relating, as I contend, to
the Parthian war? The answer is, that although this war wascarried on under the auspices of both emperors, it was only Verus
who took an active part in it, Marcus Aurelius not even visiting
the East. In the coins illustrating the episodes of this war Verus
appears alone quite frequently. The second question is: Wehave in the Capitoline reliefs untouched portrait heads of Mar-cus Aurelius, whereas we have none of Lucius Verus. Whatproof is there of any Parthian arch of Verus, to which such a
series of reliefs as those I imagine could have belonged? In the
first place the Notitia speak of an Arcus Veri on the Via Appia,which we have every reason to believe was for the Parthian tri-
umph, as I have proved that arches for eastern triumphs were
built on the Via Appia and for northern triumphs on the Flaminia.
Besides, there is a relief in the Torlonia collection belonging prob-
ably to this series, in which the emperor has always been thoughtto be Verus. Though I have seen a photograph of it, I have not
been able to examine this relief, owing to the inaccessibility of the
collection. I do not venture to assert that this relief was from
l Hist. Ill, 12, 13, 14, 31.
THE ARCH OF CONSTANTINE III 7
the Arcus Veri, but I do suggest that the Arcus' Veri may have
been despoiled to decorate the arch of Constantine.
As for the ascription to Marcus Aurelius rather than to Verus
of the figure of the emperor in the various reliefs of the attic, there
is not the slightest reason for it. The present imperial heads are
all modern, made in 1731, for Pope Clement's restoration. For
more than two, or perhaps three, centuries before that time the
emperor had in each case been headless. In all probability the
missing heads were of Constantine, and these were easily de-
tached because they had themselves taken the place of other
heads and had been loosely fastened on. These other heads
were themselves not those of the original emperor, I believe, but a
rifadmento of the latter part of the third century. This is a
point which will now be cleared up, so far as is possible.
The next point is : Wnen were these eight reliefs placed on the
attic of the arch? The matter is simple enough for those whofollow the old theory that the arch was built by and for Con-
stantine; but if the arch had been in existence since the time of
Domitian and the attic that we now see takes the place of the orig-
inal attic that was destroyed, it becomes a question whether the
attic belongs to the Constantinian restoration or to a slightly
earlier one of the third century.
In so far as the structure of the attic is concerned, it has already
been noted, that whereas the whole of the arch up to the attic is
of solid structural marble, the attic is a hollow construction in
the form of a barrel vault of rubble and brick, against which the
eight carved panels were set. A study of the construction shows
that it cannot be earlier than the last half of the third century;
its date would range approximately from 270 to 315, so far as
can be judged from the brickwork facing. It might have been
built under any emperor from Aurelian to Constantine.
The next clue is historic. The attic would be connected with
a restoration of the arch due to some triumph of an emperor pre-
vious to Constantine, or to Constantine's restoration. Dio-
cletian's triumph would be eliminated, as it was commemoratedboth by the Arcus Novus of the Via Lata and the pair of me-
morial columns in front of the Curia. The most probable occa-
sions seem to be the triumphs of Aurelian (273) and of Probus
(279).
The third clue is by far the most important. It is the head of
the praetorian prefect. There are several instances of the inten-
8 A. L. FROTHINGHAM
tion of Roman sculptors of historical reliefs to give an exact por-
trait of the emperor's chief of staff, who was next in importanceto the emperor himself in time of war. He stands close to the
emperor, usually behind him; he is with him when he is on the
raised platform. On the arch of Beneventum, beside fine por-traits of Licinius Sura and Hadrian is one of Livianus as prae-
torian prefect. Out of the eight reliefs of this attic, six have the
praefectus praetorio.
The reason for his absence in the other two is obvious. In one
case all the figures beside the emperor in the entrance scene (North
1) are ideal figures Virtus, Fortuna, Felicitas and Mars. There
is no place for mortals. In the other case the subject is a con-
giarium to the people; a civil scene in which the praetorian pre-
fect, who was a purely military functionary, took no part. His
place was taken by the praefectus urbi or praefectus annonae.
In the six reliefs where the praefectus praetorio appears, the
head is a portrait study of one and the same man, and this manis supposed by Mr. Jones to be M. Bassaeus Rufus, known to
have been the praetorian prefect of Marcus Aurelius at the time
of the Marcomannian-Sarmatian wars (168-177). But in
studying this figure even from a distance I had suspected for a
long time that the head had been recut and was not, as has alwaysbeen supposed, the original portrait. When I was able to exam-
ine the reliefs close at hand, this suspicion became a certainty.
The head had been worked over to change it from a portrait of a
prefect of Marcus Aurelius or Lucius Verus 1 to one of a prefect
of the later emperor under whom the reliefs were placed on the
attic. In only two cases was it thought necessary to change the
heads, those of the emperor himself and his prefect. The rest
of the figures were not important enough to count. In the case
of the head of the emperor himself, as it was alwa> s in the round
and at quite a distance from the background, it was easier and
better to cut off the original emperor's head and substitute an
entirely new head of the reigning emperor. This is what was
done. But these substitutes became quite easily detached and
were lost or removed before the Renaissance. It is only a con-
jecture to say that they were heads of Constantine; this is quite a
probable conjecture, however, and Constantine may have re-
placed an earlier substitute, as will become evident from what
1 The prefect of the Parthian war was L. Furius Victorinus (159-167).
THE ARCH OF CONSTANTINE III &
follows. On the reliefs in the Conservatori the original heads
both of Marcus Aurelius and his praetorian prefect remain.
The head of the prefect presented quite a different problem from,
that of the emperor when it became necessary to change it. It
was not even in high relief, so that it could not be removed, but
must be recut in situ. The rest of the figure was not touched.
I have photographed two of these heads, and they are given mFigures 7 and 8. Even a superficial glance will, I think, satisfy
any unprejudiced observer that the technique of this head differs
radically from that of every other head in any of these attic reliefs*
FIGURE 7. RECUT HEADPREFECT IN SOUTH 1
(Rex datus)
FIGURE 8. RECUT HEADPREFECT IN SOUTH 2
(Captives}
This fact was granted by every one who ascended the scaffold to-
examine the reliefs. In Figure 9, a typical group of heads show&
the technique of the original sculptor. They are in the familiar
Marcus-Aurelian style. This original style shows the extreme
use of the drill, with deep grooves at right angles to the surface
and undercutting; with curly hair and beard; strong contrasts;
dramatic expressiveness; full lips, usually parted; deep-set eyes;
fairly good modelling. If we turn to the prefect's head, we find
that the drill is used in quite a different way, not driven deep and
at right angles but diagonally; that the only deep grooves are
some that were not obliterated intherecuttingin parts usually less-
10 A. L. FROTHINGHAM
prominent and closer to the background; that there is no under-
cutting or contrast of light and shade; that the hair is cut downso as to follow, instead of concealing, the outline of the head.
The mouth also has thin lips; the moustache is almost or entirely
eliminated, and the lips tightly closed. There is little or no
modelling of facial planes, the forehead being marked with sharplines such as the original artist never used. The treatment is
crude throughout, showing a period of decadence quite unsuited
FIGURE 9. GROUP OF HEADSIN SOUTH 1 (Roman Style)
to the age of Marcus Aurelius. 1 If any one should be inclined,
nevertheless, to argue that the requirements of portraiture mighthave forced the sculptor to a flat treatment of hair, thin lips, etc.
he may be referred to the untouched head of the prefect on one of
the reliefs in the Palazzo dei Conservator!, where the techniqueis not different from that of the rest of the reliefs. The prefecthere accompanies the emperor on horseback, before whom twobarbarian chiefs are kneeling. In this untouched head there is
1 The split in the neck of Figure 8 appears to have been due to damage doneto the relief, probably in the course of taking it down from its original positionor setting it upon the arch 01 Constantine. The parts above the break are
original, not restored.
THE ARCH OF CONSTANTINE III 11
just enough resemblance to our heads to show the common
origin; how the later sculptor had to work on a face with aquiline
nose, bald forehead, and rather straggling beard. I may say, also,
that in the case of one of the attic reliefs, that of the Via Flaminia
(North 2), the original head of the prefect, almost entirely con-
cealed behind the emperor's head, appears to have seemed so
inconspicuous as to have been left almost, if not entirely, un-
touched.
Granting, therefore, that in five cases out of six the head of
the prefect was recut to resemble the prefect of the ruling em-
peror at the time of the transfer of the reliefs to the attic, does the
technique of this recutting give any indication of the date whenthis transfer was made and, if so, how does it agree with the date
indicated by the brickwork of the attic? We know that the
dramatic, contrastful style of Marcus Aurelius lasted, with ever
diminishing value, through the reign of Caracalla (f217). Thetime of Alexander Severus (222-235) seems to have been transi-
tional, with a return to delicacy of effects. Then there begins a
thin, flat, dry style, with increasing loss of technical ability and
life, with stippling often used in place of channelling, with shallow
(instead of deep) grooves, with thin lips, flat eyebrows, eyes a
fieur de t&te, hair trained flat, and beard thin and hardly changingthe contour of the chin. This style lapsed into crudity after the
time of Claudius Gothicus (268-270). Then, under Diocletian
(285-305), an abortive revival took place which continued under
Constantine. It did not pervade the entire field, but by the side
of inept and lifeless works, there are others, such as the base of
Diocletian's memorial column in the Forum and some statues of
Constantine and his family. Here we find the law of frontality
and a successful use of contrasts of light and shade and a return
to deep grooves outlining the figures against the background.In any case, as Constantine, immediately after his victory over
Maxentius in 312, abolished the praetorian guard and the office
of military praetorian prefect, it is obvious that no portrait of
such a non-existent official would have been cut in his time.
Evidently there is only one point in this evolution where the
recut prefect's head will fit into the scheme: the period after the
death of Claudius Gothicus and before the accession of Dio-
cletian, between 270 and 284. Sculpture was decadent, but it
had not yet entirely lost the ability to portray individual traits.
None of the characteristics of the styles of Diocletian and Con-
12 A. L. FROTHINGHAM
stantine are present. The period is then circumscribed to the
years of the triumphs of Aurelian and Probus. Between these
two triumphs I will not venture to decide. This question is of
minor importance. The vital point is that this head was not re-
cut in the time of Constantine. I consider that these five heads
of the prefect, recut in the time of Aurelian or Probus, may be re-
garded as a conclusive proof that the attic was rebuilt at that
time and not under Constantine. Their evidence coincides with
that of the historic probability and that of the structure of the
attic.
A great deal has been said about Germanic and Sarmatian typesand costume in connection with these reliefs, and this would mili-
tate against connecting any of them with an oriental campaign.As I cannot enter into a detailed description of the subjects in
this paper, I shall merely call attention to the fact that the use
of trousers and mantles of this type was common, as everybody
knows, to Orientals as well as to the north-Europeans of this
time. In the relief of the standard with the two imperial images,which I have considered to be the leading panel of the
" Parthian"
series, there are two barbarians. In the scene of the "Inaugura-tion of the King" there are five or more barbarians. The types,
especially in the latter relief, seem to fit an oriental race as well
as a Germanic race, or even better. A study of the heads from
this last scene (Fig. 9) will show what I mean. Of quite a dif-
ferent type are the two prisoners, one with his hands tied behind
his back, who are being roughly haled before the Emperor.These are of the north-European type: heavier of build and
shaggier of hair.
All that I have attempted to do in this paper is to give the
evidence furnished by the reliefs for dating the attic and for
deciding whether they are themselves take from one or from
more than one monument. A complete description will be
reserved for my general volume on the arches of Rome and Italy.
A. L. FROTHINGHAM.
PRINCETON, NEW JERSEY.
Archaeological institute of America
AMERICAN
JOURNAL OF ARCHAEOLOGY
THE JOURNAL OF THE ARCHAEOLOGICAL
INSTITUTE OF AMERICA
VOLUME XIX
1915
CONCORD, N. H.
PUBLISHED FOR THE INSTITUTE BY
Wbt !\umforb $re^NEW YORK: THE MACMILLAN COMPANY
LONDON: MACMILLAN & CO., LTD.
Srcfjaeologtcal
of America
WHO BUILT THE ARCH OF CONSTANTINE?
IV. THE EIGHT MEDALLIONS or DOMITIAN 1
IN THE papers thus far published on the arch, after the intro-
ductory paper in which I sought to prove that it was built byDomitian (87-96 A.D.), I studied some of the sculptures that were
added to the original arch during the third and fourth centuries;
either by cutting them in the Domitianic masonry, as was the'
case with the triumphal frieze across the east and west ends, or
by transferring them bodily to the arch from some other struc-
ture, as was the case with the attic reliefs and the friezes on the
north facade, or else by carving them expressly for the arch and
inserting them, as in the case of two medallions of the east andwest ends, and the spandrel figures.
In the present paper I shall attack the even more fundamental
subject of sculptures which I attribute to the original decoration
of the arch, and therefore to the reign of Domitian himself.
While my general thesis cannot be said to stand or fall by this
test, it will be greatly strengthened if I am able to show that
sculptures generally conceded to belong, by their artistic qualities,
to the time of Domitian, are so related to the structure of the
arch as to make it seem almost or entirely certain that theyformed part of the original construction; and also that they and
their surroundings were afterward modified during the changesto which the arch was subjected before and during the time of
Constantine, in such a way as to add to the probability that theywere there already, before these changes took place.
Such sculptures I believe to be the eight medallions of the
two main fagades, and the four keystones of the minor arcades,
only one of which, however, has escaped more or less completedestruction.
The eight medallions are in two groups: those in Figure 1
are the group on the south fagade; in Figure 2 are those on the
iSee A.J.A. XVI, 1912, pp. 368 ff.f XVII, 1913, pp. 487 ff., and XIX,
1915, pp. 1 ff.
American Journal of Archaeology, Second Series. Journal of the
Archaeological Institute of America, Vol. XIX (1915) No. 4.
368 A. L. FROTHINGHAM
north facade. Any apparent differences are due to the different
weathering and light. Their average diameter is ca. 2.35 m.,
and they are carved in a single slab of marble. They are groupedin pairs over each of the minor openings of the arch in a panel;
Figure 3 will show how they were connected with the decorative
scheme.
FIGURE 1. MEDALLIONS ON THE SOUTH FACADE OF THE ARCH OF CON-STANTINE (from Mrs. Strong, Roman Sculpture)
These eight medallions have attracted more attention and
admiration than any of the other reliefs, not only on account
of their artistic merit, but from the mystery that surrounds them.
There is a unity of theme running through all of them; it is the
hunting exploits of some emperor. Who this emperor was
cannot be proved, but I believe him to 'be Domitian. The
THE ARCH OF CONSTANTINE IV 369
scenes were so harmlessly impersonal that it was not felt neces-
sary to destroy them when the arch was dismantled after Domi-
tian's death. The changing of one of the imperial heads into a
portrait of another emperor, conjectured by different critics
to be Claudius Gothicus or Carus or Philip,1 and of another
original imperial head into one supposed to be Carinus or Con-
FIGURE 2. MEDALLIONS ON THE NORTH FACADE OF THE ARCH OF CON-STANTINE (from Mrs. Strong, Roman Sculpture)
stantius Chlorus,2 has puzzled recent critics who vary in their
1 This is the Sacrifice to Hercules (N. 4), where the imperial head is so
evidently recut away from the sacrificial headdress.2 The Sacrifice to Apollo (N. 2), which is not really a sacrifice, as the head
is not veiled. It is probable that the original head was veiled and that in the
recutting the drapery was cut away.
370 A. L. FROTHINGHAM
identifications. Another interesting change of the same class
is the complete substitution (not recutting) of a head of Con-
stantine for that of the original emperor.1 In all four of these
south medallions the emperor's head is enclosed in a circular
nimbus,2 a rare official Roman case of the recognition of the Sun-
cult as the religion of the state, analogous to the recognition of
the Mithraic cult by Diocletian in the base of his Memorial
Column to which I recently called attention in this JOURNAL
(XVIII, 1914, pp. 146 ff.), where Mithra himself has the nimbus,
not, in this case, a simple but a rayed circle. Recent studies
have made it abundantly clear that, for several years after the
time when the Arch of Constantine was remodelled and dedicated
to him, Constantine was officially regarded as a Sun worshipper.
The imperial heads in the four medallions on the opposite
or south fagade present quite a different puzzle. Only in one
case is the head at all preserved in the Sacrifice to Diana (S. 4).
Here there is no fracture at the neck. The head is original.
Neither is there any trace of recutting in third century tech-
nique. Yet it is difficult to assert that the features are Domi-
tianic, because they have been so badly obliterated by fire. In
fact Miss Bieber believes the face recut into a portrait of Con-
sta'ntine! In" the Bear Hunt (S. 3) the entire face has been
calcined away, but the neck shows again that there was no sub-
stitution; though whether there was recutting we have absolutely
1 The Boar Hunt (N. 1) and the Lion Hunt (N. 3). In both cases the break
at the neck is quite clear.
2 It is supposed that the nimbus was added by Constantine; and there are
still critics who believe it to be due to his conversion to Christianity; and
for this reason Philip the Arab is identified with one of the other heads because
there is a tradition that he had secretly become a Christian. There is really
no foundation for such a fantastic notion. It arises from the quite general
ignorance as to the history of the nimbus in pre-Christian and non-Christian
spheres. To keep in the sphere of Roman imperialism, there is a notable
example of a medallion of the Emperor Diocletian with a nimbus of this same
pattern. Now, no one can accuse Diocletian of being a Christian! His
nimbus is a sign of his adoption of the solar cult of Mithra in the same wayas its use in the case of Constantine is a sign of his worship of the Sun god
Apollo. I am preparing material for the history of the nimbus or sun-glory,
and this includes examples of practically every century of Greco-Romanart from the fourth century B.C. to the middle Roman Empire. The Chris-
tian nimbus was a pure case of the adoption of a perfectly well-known "pagan"symbol, and was due to the association often expressed of Christ with the
divine Sun. It was plagiarism, pure and simple. During the third century
A.D., the prevailing worship was Sun worship.
THE ARCH OF CONSTANTINE IV 371
no means of knowing. In the Departure for the Chase (S. 1),
the loss of the head is absolute, also apparently by fire, primarily,
and by fracture, secondarily. Finally, in the Sacrifice to Sil-
vanus, besides extensive damage by fire there was a violent
fracture, apparently due to the impact of some heavy object
falling from above, which split off the whole front of the torso
as well as the head. Evidently, then, the current opinion that
the four imperial heads on this face were left intact is unprovablein at least two cases, and may be considered only as a probability.
It is only on stylistic grounds that they can, therefore, be
ascribed to Domitian; except that in the one case where the
original head seems untampered with and merely injured the
Sacrifice to Diana the evidently delicate and beardless face
belongs apparently to an early emperor of the Domitianic type,
which would exclude almost any other possibility.
Older critics had attributed the medallions to the reign of
Trajan. With the progress made during the last two decades
in a critical knowledge of Roman sculpture, two opinions were
brought forward, almost simultaneously: one attributing themto Hadrian and the other to Domitian. The majority of critics
appear to have adopted the Domitianic theory, which seems to
be convincing. Of course it also fits perfectly into my theoryof the arch and its decorative history. But, it will not be neces-
sary for me to repeat here the arguments on the Domitianic side. 1
1 They may be seen in brief in Mrs. Strong's Roman Sculpture. The pro-
mulgator of the Hadrianic theory was Paul Arndt in Bruckmann's Denkmdler,
pis. 555, 559, 560, 565, published in 1903. The Domitianic theory was pre-
sented by Stuart Jones in Papers of the British School at Rome, III, pp. 216-
271, from a study of the medallions in 1904. This was a great advance on
Petersen's monograph in Rom. Mitt. IV, 1889, pp. 314-339, where the Tra-
janic date is still unquestioned. There is a symposium of opinions of Sieve-
king, Studniczka, Reinach, Esperandieu, S. de Ricci and Bieber in Revue
Archeologigue, XV, 1910, pp. 118-129 (with fine cuts of the heads), p. 170;
1911, p. 465. Sieveking published a detailed study in Rom. Mitt. 1907,
pp. 365 ff., in which he made a difference between the medallions on the north
side, which he thinks Hadrianic and those on the south which he believes
to be Domitianic. This theory was opposed by Miss Bieber in Rom. Mitt.,
1911, pp. 214 ff., 'Die Medaillons am Konstantinbogen.' She shows how
apparent differences in style between the south and north medallions are only
apparent and due to different weathering and to different effects of light and
shade which affected the photographs on which Sieveking largely based his
opinion. Sieveking's partial retractation appeared in Berl. phil. Woch. 1911,
No. 39. To this explanation of apparent differences I would add the greater
damage inflicted on the south medallions by fire.
372 A. L. FROTHINGHAM
This paper will deal with the reliefs entirely from the struc-
tural point of view, without discussion of questions of style or
subject. I make a slight exception in the next few remarks
because of a bearing on the questions- discussed.
The first exception relates to the heads of emperors already
spoken of as changed by substitution or recutting, on the four
north medallions: two of the new heads being of Constantine and
the other two of two different emperors who cannot be identified
with certainty, but are usually thought to belong to the second
half of the third century. All four have the solar nimbus.
What explanation of these changes can possibly be given on the
basis of the old theory that the arch was built by Constantine?
The only plausible suggestion that I have noticed makes Con-
stantine the author of all these changes. That he had onlytwo of his own heads used and that for the other two he made
portraits of two earlier emperors is supposed to be due to his
desire to do honor to and assert family connection with previousFlavii and imperial solar worshippers. Constantine's historians
claimed Claudius Gothicus as an ancestor of Constantine, and
tried to connect him with the earliest Flavii. But there is a fatal
flaw in this argument. Its authors seem unaware of the fact
that the heads of these two other emperors whether they are
Philip, Carus, Carinus or some others are executed in a tech-
nique simply impossible in the time of Constantine. This
technique, which was current only between ca. 230 and 275
A.D., was characterized by abuse of"stippling," flat and thin
hair, eyesdfleur de tete and other peculiarities which I have noted
in a previous paper. It is absolutely distinct from the workdone in the time of Diocletian and Constantine.
The real explanation of the changed heads seems to me to be
simply this: When the arch was associated during the course
of the third century with the triumphs of different emperors,some sculptures were added or changed to record each triumph.In another paper I shall study the eight half figures of emperorscrowned by Victories which were inserted in the masonry of the
two minor arcades to celebrate their triumphs. These figures
are in the various manners of third century sculpture. Appar-
ently it is two of these emperors whose heads were cut on the
medallions, probably at the same time that their triumphant
figures were inserted in the arcades below.
I merely refer to this detail of the medallions in order to show
THE ARCH OF CONSTANTINE IV 373
that it favors my contention that they formed part of the arch
as early as the third century and before, and were modified in
situ to adjust them to the decorative features that were added
and to the remodelling of parts of the arch surface. A moredetailed discussion of the heads will be in order in the future
paper on the eight imperial portraits of the minor arcades.
Just one more remark before entering upon the constructive
discussion. The injury to many of the medallion heads with the
greatest projection has already been discussed, as being due
partly to fire and partly to the impact of something heavy falling
from above. For instance, in the"Sacrifice to Apollo," the head
of Apollo and that of the emperor's attendant with the horse
were both broken off; recovered and reattached. These heads
projected sufficiently (10 to 18 cm.) not to be protected by the
frame. I have not questioned the current opinion that the em-
peror's head is here the original head worked over. But the
point is debatable. There is an evident break at the neck; a
magnifying glass will, I think, show traces of it even in Figure2. It may easily be argued that the whole head was done in
the time of the emperor whose portrait it is; or, as in the twoother cases, the head may have been reattached immediatelyafter the damage done at Domitian's death, and then recut in
the third century. In any case, the fractures in these andother figures of the medallions, especially where they are diagonalor almost vertical, distinctly favor my contention that the medal-
lions were damaged by the bronze figures and groups, the marble
slabs and statues that were cast down from the attic after the
death of Domitian.
Entering now into the heart of my argument, there are five
main characteristics that bear on the structural relation of the
medallions to the arch: (1) their shape; (2) the shaving of the
lower curve of several of them; (3) the closeness of the joints;
(4) the treatment of the surface around them; (5) the marble
veneer of the enclosing panel.
I. Shape of the Medallions. It has been taken for granted that
the single slab on which each medallion is carved was circular
and corresponded exactly to the outline of the enclosing frame.
Both of these suppositions are wrong. The frame, in the first
place, is not a perfect circle. It is so for about three-quarters of
its circumference, but the other quarter, corresponding to its
base, has been given a slight flattening, a depressed curve which
374 A. L. FROTHINGHAM
is so carefully graduated as to have remained unnoticed. In
the Boar Hunt scene on the left end of the north side (N. 1) this
blunting amounts to four centimetres, as the width of 2.39 m. is
reduced in the height to 2.35 m. It must be remembered that
the four centimetres are taken from about a quarter of the cir-
cumference.
Far more important than this is another fact of which I be-
came aware only when I studied the medallions very closely
from the scaffold. It is that this circular-appearing slab rests
on a square base; that from a quarter to a fifth of the lower cir-
cumference of its frame is not fitted into a slab cut with a curved
edge to receive it, as is the case with all the rest of the circum-
ference, but that what appears to be a separate base block is in
reality an integral part of the medallion. The true outline of
FIGURE 3. SACRIFICE TO APOLLO (N. 2)
(A shows base-line of figures (a) ;actual outline of cornice (6) ; perfect circle
(c); and squared base (d). B shows the plain block outline)
the medallion block is given in Figure 3, where in (A) the dotted
line stands for the true circle at the base, and the black line for
the outline of the actual medallion cornice, while in (B) is the
actual shape of the block in which the medallion is cut.
The architects to whom I communicated my discovery of the
square base of the medallion blocks were of the opinion that it
was a strong argument in favor of my theory that the medallions
were an original part of a Domitianic arch, especially in view
of the extreme closeness of the joints. In confirmation I oughtto call attention to the fact that these medallion blocks are not
thin slabs that could easily be transferred. I tried to gauge
THE ARCH OF CONSTANTINE IV 375
their thickness, but found it impossible to run into the joints
the thin steel skewer which I had brought for the purpose. Onlya long slender hat-pin could be inserted ! I ran it to its end with-
out meeting any obstacle so I know that the blocks are more
than 35 centimetres thick; how much more I cannot say.
II. The Cut in the Base of some Medallions. The difference
between the slight flattening of the lower part of the circle on the
north side in medallions 1-3, which was evidently planned at
the outset by a keen-eyed and truly aesthetic artist, and the
more obvious and crude flattening on the south side has never
been noticed, much less explained.
Evidently there must originally have been the same slight
flattening as on the north. But, for some reason this original
outline was modified at some time. The outline was flattened
at least twice as much, in a crude fashion. When this was done
FIGURE 4. SACRIFICE TO SILVANUS (S. 2); TRIMMED BASE
the encircling frame-moulding must have been so seriously en-
croached upon that at the extreme base it must have been prac-
tically obliterated. It was consequently seen by the stone-
cutters that in order to conceal this defect it would be necessary
to shave off the face of as much of the lower part of the medallion
frame as was affected by this change of outline. In this way the
outline of the frame was totally done away with. The tooling
of the new surface is very rough and shows its late date, in a
period of decadence. In the last medallion on the right there
was evidently a slight variation in level, because it was found
necessary to trim the bottom but very slightly. The "Sacrifice
to Silvanus" (S. 2) in Figure 4 is typical of the three south medal-
lions 1-2-3 in the amount of the trimming.In order to make the whole matter perfectly clear, I give in
Figure 5 a diagram based on the"Sacrifice to Silvanus." It
376 A. L. FROTHINGHAM
shows how the vertical line of the square medallion base cor-
responds to the horizontal base line of the relief and that this
point of intersection is the point where the designer broke the
line of the circle in order, apparently, to give an appearance of
greater stability to his base. This is represented by the dotted
line. Then, when the frieze was inserted, the base was trimmed
crudely to the present line not as evenly as in the cut makinga small gap between medallion and frieze.
FIGURE 5. SACRIFICE TO SILVANUS (S. 2); SHOWING VARIATIONS
What was the reason for this mutilation? It seems to me
perfectly evident. It constitutes, in fact, one of the most cogentreasons for believing the medallions to be part of a Flavian arch.
One has only to note that if this change had not been made there
would have been absolutely no space between the medallion
frames and the friezes below them! Whether the friezes were
inserted or were carved in the existing masonry makes no prac-tical difference. At whatever time the frieze was carved or
inserted it became absolutely necessary to cut away part of these
THE ARCH OF CONSTANTINE IV 377
medallion frames in order to continue without interruption the
panelling of colored marble veneer around the entire medallion.
To a lesser degree the same process was required on one onlyof the north medallions: the one on the extreme right, the
"Sac-
rifice to Hercules." Here the flattening needed was considerably
less, but is quite evident and accompanied by a similar but less
extensive cutting away of the frame and lower surface. 1
III. Close Joints, Surface Tooling and Veneer. If the medal-
lions had been taken from an older monument and been built
into the arch in the course of construction in the time of Con-
stantine, they would have been fitted into a wall built of already
prepared blocks, which would have been set with their faces
sufficiently in retreat to allow of receiving a marble veneer
without the need to cut them back again. The architect would
in that case have been perfectly free to place them at any height
and in any relation to the other decorative features that he
chose. He would have arranged the sculptured friezes under
the medallions in such a way that each should not interfere with
the other, but they should have an adequate space between themfor the enclosing marble veneer.
Now, it is perfectly evident that things did not happen in this
way. We have already seen that the architect could not have
been free to correlate as he pleased the friezes and medallions, but
was obliged to mutilate half of the medallions to get room for
the friezes. Also a glance at Figure 7 will show that the use of a
veneer could not have been present in the mind of the architect
when the medallions were set in the arch because the blocks
were refaced and cut back in situ, when the medallions were
already in place, if we can trust the evidence that stares at us
from the panels.
There is a great difference between the south and north sides
in the condition of the surface of the panels. On the north wecan study it as it was originally recut, some time in the third
century; but on the south side it is impossible to be at all sure of
the date of the present surfacing, which is a pot-pourri of roughand smooth masonry, with occasional wide joints. The masonryof the north side shows extraordinarily close joints everywhere.
1 In the medallion to the left the Lion Hunt it is not easy to say whether
there may not be a slight blunting because, as the body of the lion both inter-
rupts the frame in any case and also gives a naturally irregular outline, the
question is an open one.
378 A. L. FROTHINGHAM
Moreover, the courses are formed, at the base, of vertical, not
of horizontal blocks, giving much wider courses, and showinghow the designer planned the masonry with due regard for the
insertion of the medallions at this level. In connection with the
closeness of the joints, I would call attention to the contrast
with the loose joints on the upper line of the friezes, as a further
indication that these friezes were inserted.
The next point is the tooling of the surface of the panels. I
am inclined to believe that these panels were not formed around
the medallions at the time of the first restoration of the arch;
that is to say when the columns were added, the main cornice
restored, and the resaults added to it above the columns. It
seems as if when the pilaster responds to these columns were set
into the arch the original surface still existed. Of course, the
theory that I .hold of successive additions during the third cen-
tury in connection with the triumphs of different emperors, in-
volves a different date for, let us say, the insertion of each pair
of friezes. The two north friezes were set in, I believe, at a
different time from the two south friezes; and the triumphalfriezes on the two short east and west ends, with their short
returns on the north and south fronts, belong to a third and still
later date, as they presuppose the existence of the other friezes. 1
I shall not enter into this question here, and mention it merelybecause it was necessary to say this in order to explain that the
panelling of the north face was done at a different time from that
of the south face (probably earlier) and this may explain the
difference in technique.I do not think that, on reflection, any critic would contend
that the use of a veneer in these panels was part of the original
plan. There are two reasons not already mentioned. The first
is that where a marble facing is found it is set against a core of
rubble, of brick, or, in earlier times, of travertine. But in this
arch the structure itself is of marble blocks, so that marble uponmarble is like painting the lily. The application of a marble
veneer to a marble structure must therefore be due to circum-
stances arising after the erection of the structure. The second
1 If there had been no friezes already on the main faces of the arch when the
triumphal procession was cut into its surface, there is no doubt in my mindthat this procession would have started on the left end of the north face and
occupied the spaces over the minor arcades. It is quite abnormal that it
should be split up as it is; it is so by force majeure, owing to pre-existing
circumstances.
THE ARCH OF CONSTANTINE IV 379
reason is the special nature of the facing. The rule is that where
there is a marble facing to a structure it consists of more or less
heavy slabs. Only in a few early Augustan structures such as
the Porta Praetoria at Aosta and the city gates of Spello was
there a use of thinner slabs; but even in these cases the marble
FIGURE 6. ARCH OF CONSTANTINE FROM THE NORTHEAST
is considerably thicker than the veneers on the arch of Constan-
tine and was not colored. In fact, the technique of these ve-
neered panels is a unique example of the transfer to the exterior
of a structure of a process elsewhere used only for interior wall
decoration. We are familiar with the use of colored marble
veneers in the halls of Roman basilicas, thermae, palaces, etc.,
and with the transmission of this brilliant and permanent form
380 A. L. FROTHINGHAM
of interior decoration to Byzantine religious and civil art. Thefact that only on the arch of Constantine, if I am not mistaken, is
it used on an outside wall, is a further proof that it was not an
FIGURE 7. ARCH OF CONSTANTINE; SOUTH FACADE, WESTERN PANEL
original but an emergency method, to solve a difficult problemof re-surfacing.
Before attacking the details of the treatment of the panel sur-
THE ARCH OF CONSTANTINE IV 381
faces, the question of
the surface of the arch
masonry must be at
least glanced at in so
far as it affects these
problems. The best
view for this purposeis that given in Figure
6 looking diagonally
from Northeast to
Southwest; that is,
across the north
facade from the east'
end. This shows the
original Domitianic
masonry, practically
untouched, of the
whole short end, up to
the main cornice and
around the corner of
the main face as far as
the pilasters. It is
plain how the trium-
phal frieze was cut
into two courses of the
old blocks in the third
century and broughtaround the corners as
far as the pilasters.
Then, beyond the
pilasters, the surface
was cut away, a little
earlier in the third
century, around the
pairs of medallions, in
order to connect the
medallions with the
newly inserted friezes,
and leaving thesemedallions as an oasis
in a desert of thirdFIGURE 8. SACRIFICE TO APOLLO (N. 2);
TOOLING MARKS
382 A. L. FROTHINGHAM
century work. For, in the centre, even the archivolts of the
main arcade were recut. The architectural part of this trans-
formation will be treated in another paper.
At present I shall merely say that the surface around the medal-
lions was cut back between 1 1 and 12 centimetres, and that, after
architrave above and sculptured frieze below had been used to
frame the panels at top and bottom and pilasters at the sides, a
coat of cement from 4 to 5 centimetres thick was spread over the
ground of the panels and against this were set thin veneer slabs
of various colored marbles, fastened also by lead, and forming a
brilliant ground for the medallion reliefs. The veneering slabs
varied in thickness from 0.75 to 1.5 centimetres. We cannot
say what was the color scheme of the veneer, because it was
almost entirely torn away during the Middle Ages. In prizing
FIGURE 9. TOOLING AT BASE OP BOAR HUNT AND SACRIFICETO APOLLO (N. 1-2)
off the veneer a few fragments at the narrowest points were left
too small and few in number to tell much of a story. Manybunches of the cement backing remain. At intervals the square
holes for the lead can be noted.
Except for the places where some cement remains we can studythe tooling of the marble blocks, which is of course rough and
intended to be concealed. Now, the technique of this tooling is
interesting. In fact I believe it to furnish the strongest of all
the arguments in favor of my theory that when the surface was
cut away the medallions were already a part of the arch.
The tooling marks can best be studied on the north face, where
the treatment, as I have alraady said, is quite different from that
of the south face. There are two important peculiarities of
this tooling: (1) that it is often continuous from one slab to
THE ARCH OF CONSTANTINE IV 383
another across the joints; and (2) that its lines are extremely
irregular and diagonal, being evidently determined in their direc-
tion, their changes, and juxtapositions by the previous presenceof the medallions.
The facts can best be studied in the panel which contains the
Boar Hunt and the Sacrifice to Apollo (N. 1-2), illustrated in
Figures 7, 8, and 9. A piece of veneer remains at both top andbottom of N. 1, at the bottom of N. 2, and between them in the
centre. The top piece is 12 cm. wide; the bottom pieces 7 cm.
wide; the piece between the medallions is 16 cm. wide. All but
the latter, which is porphyry, seem to be white marble. The
porphyry fragment is 0.75 cm. thick, is set 5 cm. away from the
surface of the masonry*and 5.5 cm. back of the medallion frame.
The block forming the medallion base is very roughly and
irregularly cut away near the frame of the medallion, so that the
FIGURE 10. BASE OF SACRIFICE TO HERCULES (N. 4)
surface is not flat but curves concavely, especially at the narrower
part of the neck, as if the work had been done after the frieze
had been put in position.
The inference to be drawn from the direction and length of
the tooling lines is extraordinarily clear. While this is com-
paratively evident in the photographic illustrations of Figures7 and 8, I have made it plainer by a careful linear facsimile in
Figure 9. The three characteristics I have already mentionedare here: The continuation of the lines across the joints; the
fact that the tooling does not, as would be natural, follow the
rectangular lines of the blocks; but that its lines are varied so
as to show that they were conditioned by the medallion frames.
In Figure 10 is the lower part of N. 4 (Sacrifice to Hercules)where it is very plain how crudely the bottom was shaved awayto make room for the frieze, which would otherwise have come
directly against it. The light shining on the left curve of the
384 A. L. FROTHINGHAM
base shows how there is no joint below the point where it inter-
sects the last vertical joint; that is, how the slab rises on a squarebase.
The reason then for considering the medallions an original
part of the arch are:
(1) That they are of Domitianic art, and that the arch is Dom-itianic
;
(2) That they are innocuous and generic in theme and so
could be spared by the iconoclasts at Domitian's death;
(3) That they set in the masonry of the arch with perfect joints
of the best period and with a square base that forms part of the
course masonry;
(4) That they were already in the masonry when its face wascut back all around them to add the marble veneer, as is shown
by the tooling;
(5) That they were already there when the friezes below were
carved or set in, as is shown by the way it became necessary to
mutilate their base line in order to leave any space between themand the frieze;
(6) That the changes in the imperial heads harmonize with
the idea that the arch was used to commemorate triumphs of
emperors of the third century before the time of Constantine.
A. L. FROTHINGHAM.PRINCETON, NEW JERSEY.
Recommended